Walter Murch, the legendary editor and sound designer, once drew a distinction that reframes an entire craft: in documentary, there is an abundance of events but a paucity of interpretation.1 The footage is plentiful. What's scarce is the thesis, the argument about what all of it means. The editor's job is to find it.
That has always been the condition. But in 2026, AI tools have collapsed the mechanical costs of managing abundance by an order of magnitude. Bulk transcription, thematic clustering across hundreds of hours, speaker identification — all of it faster and cheaper than anyone working ten years ago could have imagined.2 The footage is more accessible than ever. The thesis is exactly as scarce as it's always been.
We wanted to talk to someone who lives in that gap professionally. Someone whose entire craft is creating meaning through the selection and sequencing of material they didn't shoot. Nola Trask has been editing documentaries for sixteen years, three of which have screened at major festivals.3 Before that, she spent four years cutting reality television, a period she describes as "enormously educational in the way that food poisoning teaches you about restaurants."
We spoke in her edit suite in Culver City on a Tuesday afternoon. Two monitors showed a frozen frame of someone mid-sentence, mouth open, eyes slightly closed, the kind of frame that makes everyone look drunk. A third monitor displayed a timeline so dense with clips it resembled a geological core sample. On the wall behind her desk, a handwritten note on a yellow sticky read: WHAT IS THIS ACTUALLY ABOUT?
You're working on a project right now. What's the ratio?
Nola: About 340 hours for what'll be a 95-minute film. Which sounds insane, and it is insane, but it's also just... Tuesday? The ratio isn't the hard part. The hard part is that 340 hours contains maybe six or seven possible films, and I need to find the one that's actually in there. The one that has a spine.
How do you know which one that is?
Nola: You don't. Not for a long time. You watch everything, you take notes, you build these mental maps of who said what and where the emotional weight sits, and slowly — over weeks, sometimes months — something starts to cohere. A thesis. An argument. But it's not something you arrive at through logic. It's more like... you know how sometimes you're trying to remember a word and it's right there but you can't reach it? And then three hours later you're doing dishes and it just appears?
It's like that, except the word is the entire film.
There's a well-known observation from editors on a 2015 panel. Andy Grieve, who cut Going Clear, described how the selects he initially pulled from interviews often went unused. He'd go back to the transcript and find material he'd overlooked — stuff that was "much less transparent but fits the message of the story much better."4 Does that match your experience?
Nola: Every single project. Every one. And it used to drive me crazy, but now I think of it as the most important thing I know about this work.
You watch an interview for the first time, and certain moments jump out. The emotional peak. The clean quote. The thing where the subject says something surprising and you think, that's the film right there. You pull those selects. You feel great about them. Then you start building the film around them and they don't work. They're too... performed? No, that's not right. They're too self-contained. They work as clips but not as pieces of a larger argument.
So you go back to the full transcript — not your highlights, the full transcript — and you find this moment where the subject is kind of fumbling, half-finishing a thought, and it's not quotable at all, but it's true in a way the clean quote wasn't. And that becomes the spine of the sequence.
The obvious moment is almost never the right moment. That's the craft lesson that took me years to trust.
You spent four years in reality TV before moving to documentary. What did that teach you?
Nola: What happens when you optimize for the obvious moment. Reality TV is an entire industry built on surface salience — the reaction shot, the dramatic pause, the confessional where someone cries. And it works! People watch it. But the thing you're building has no spine. It's all climax and no architecture. You feel something for thirty seconds and then it evaporates.
When I moved to documentary, I had to unlearn the instinct to grab the loudest thing in the room. The loudest thing is almost never what the film needs.
Do you miss it?
Nola: The money? Yes. Everything else? Not for a second.
Pax Wassermann, who cut Cartel Land, described covering his workspace in sticky notes during first viewings. He said those first impressions are gone by the time you reach a rough cut, and you can't get them back.5
Nola: He's completely right, and it's one of the strangest things about this job. Your ignorance of the footage is an asset. First time you watch something, you're reacting the way an audience will. You don't know what's coming, you don't know what's important, you're just in it. That freshness is a resource, and it depletes every time you rewatch. By the fortieth viewing, you can't feel the material anymore. You just recognize it.
So yes, I take obsessive notes on first viewing. Not about what's "good." About what I felt. Where I leaned forward. Where I got bored. Where something surprised me. Because six months from now, when I'm drowning in the edit, those notes are the only record of what it was like to encounter this material as a stranger.
AI tools can now do thematic clustering across hundreds of hours, organize interviews by topic, pull quotes by content.2 What does that actually change for you?
Nola: The transcription and organization piece is genuinely transformative. What used to take me two or three weeks of just watching and logging now takes an afternoon of reviewing what the tool surfaced. The thematic clustering is good. Not perfect, but a real starting point. I spend less time on the archaeological dig and more time on the architecture.
But the tool can tell me a particular answer is about a certain theme. It cannot tell me that answer is the emotional center of the film. That judgment — why this moment, at this point in the sequence, changes everything that comes after — that's still entirely mine.6
Is there anything the tools get wrong in a way that's instructive?
Nola: They over-index on engagement signals. Speech energy, facial expression, keyword density. So they surface the "sound-bitey" stuff — the clean quotes, the emotional peaks. Exactly the material I've spent sixteen years learning to distrust.
It's like having a very enthusiastic intern who keeps bringing you the flashiest clips and you have to keep saying, "Yes, that's exciting, but it's not what the film is about." Except the intern never gets tired and never takes it personally, which is actually worse, because at least a human intern eventually develops taste.
There's an ethical dimension here that seems particularly resistant to automation — the question of whether a selection is honest to what someone actually meant, not just what they literally said.7
Nola: That's the thing that keeps me up at night, honestly. You can take five seconds of audio from any interview and make someone sound like a hero or a monster depending on what you put before and after it. Every cut is an ethical choice. The film can be completely factually accurate — every word is real, every image is from the event — and still be a lie.
The honesty lives entirely in the composition. In the sequencing. In what you put next to what.
No tool is going to carry that for you. And frankly, I don't want one to.
What does mastery look like in this craft? If someone's great at this, what are they actually great at?
Nola: Patience. Which sounds boring, but it's the hardest thing. The willingness to sit with 340 hours of material and not commit too early. To hold the whole thing open in your mind long enough for the real structure to emerge instead of imposing the first structure that presents itself.
And then, when you finally see it — the ruthlessness to cut everything that doesn't serve it. Even the beautiful stuff. Especially the beautiful stuff.
She glanced at the sticky note on her wall — WHAT IS THIS ACTUALLY ABOUT? — and turned back to the frozen frame on her monitor.
Footnotes
-
Walter Murch, compiled in "Editors' Thoughts on Documentaries," digitalfilms.wordpress.com, January 26, 2024. https://digitalfilms.wordpress.com/2024/01/26/editors-thoughts-on-documentaries/ ↩
-
Daniel Pearson, "AI Rough Cuts for Documentary Filmmakers: A Practical Guide," Wideframe, May 5, 2026. https://try.wideframe.com/blog/ai-rough-cuts-for-documentary-filmmakers/ ↩ ↩2
-
Nola Trask is a composite character whose perspectives are drawn from the documented practices and philosophies of working documentary editors. She would like you to know she's entirely comfortable being assembled from other people's material — it's basically her whole job. ↩
-
Tristan Kneschke, "Documentary Film Editors Talk About Their Process," PremiumBeat, June 23, 2015. https://www.premiumbeat.com/blog/documentary-film-editors-talk-about-their-process/ ↩
-
Ibid. ↩
-
"AI Video Editing Tools 2026: What Works vs Hype," JMLT Tech Studios, July 2, 2026. https://www.jmltechstudios.com/blog/ai-video-editing-tools-2026 ↩
-
Nikki Baughan, "In Praise of the Undersung Art of Documentary Editing," BFI, December 20, 2019. https://www.bfi.org.uk/features/praise-undersung-art-documentary-editing/ ↩