Last month I was editing a three-minute short film. Spent forty minutes just staring at a render progress bar, doom-scrolling on my phone. Three years ago? Totally normal. Now I kept thinking — I could've typed a few lines into an AI video generator, waited ten seconds, and had something roughly in the ballpark.
I'd be lying if I said that didn't mess with me.
I'm talking about Sora, Kling, Runway — the whole wave. When these AI video tools first dropped in 2024, people in the industry were smug. "Look, six fingers." "Look, the water's flowing upward." Fair enough. But by this year, those jokes got real hard to make. The image quality, camera movement, lighting logic — the progress has been honestly kind of rude. A buddy of mine who's been doing post-production for eight years used an AI video maker to deliver a pitch reel to a client last month. The client watched it and said, "Looks great. Can we trim the budget a little?" He told me the story with a smile. I could hear something else underneath it.
But here's the thing — I don't really want to rehash the "will AI replace us" debate. We've beaten that horse into paste.
What actually interests me is what happens to video as a medium once the barrier to making it just... evaporates.
It used to be that making a video meant you understood composition, lighting, pacing. You probably needed decent gear. That skill tree took two, three years minimum to climb. So video creation was, by design, a gatekept thing. A "professional" thing. Now, someone who has never touched a camera can hold an image in their head and basically speak it into existence. I saw a retired middle-school Chinese teacher make a short AI-generated video about the small county town he grew up in during the '80s. The visuals were rough around the edges. But the atmosphere — that damp, coal-smoke heaviness — was more precise than a lot of stuff I've seen out of film school.
This isn't a tech upgrade. It's a redistribution of who gets to say something.
And yeah, that's also where it gets tricky. When everyone can generate a video, the act of generating one isn't worth much anymore. What becomes valuable is what you're actually trying to say. Your taste. Your instinct for whether a shot should hold for two seconds or three. The dumber the tool gets, the rarer good judgment becomes. Not a new idea, honestly. When photography showed up, portrait painters panicked the exact same way. Painting didn't die. It just became a different thing.
My own workflow has already shifted. Rough cuts, B-roll, transitional footage — if an AI video tool can handle it, I'm not doing it by hand. But the key shots, the ones where the emotion has to land on the exact frame, where the rhythm is off by half a beat and it just doesn't work — those I still do myself. Not because the AI can't. Because I need to be in there, wrestling with the footage. That struggle is part of the work. Hand the whole thing to a machine and I genuinely don't know what I'm doing anymore.
There's a line I keep coming back to: AI won't eliminate creators, but it will eliminate people who were only ever executing someone else's vision. If your entire value is translating someone else's idea into images — yeah, that's shaky ground. But if you're the one with the idea in the first place, AI just made you faster. Cheaper. That's it.
My guess for the next two or three years: short-form content, ads, product demos — the "functional" side of video — will get swallowed up by AI video generators at scale. But the stuff with a real authorial fingerprint — documentaries, experimental work, anything that carries the specific body temperature of one human being — that actually gets more precious because it's scarcer. Handmade leather goods didn't vanish because of assembly lines. They just moved to a different shelf.
I'm not pessimistic. But I can't pretend everything is the same as before.
That afternoon of scrolling my phone while waiting for a render? That's gone. And I don't think it's coming back.
