Hengdian World Studios is empty. The busiest season of the year, and actors are packing their bags. One veteran who spent a decade in the world’s largest film set just posted: ‘I’ve only worked 6 days this month. I’m leaving.’
You’ve probably heard the story before — AI is coming for creative jobs. But what’s happening in China’s short drama industry isn’t a slow replacement. It’s a silent collapse that most observers completely misread.
Here’s the number that matters: a three-minute AI-generated episode now costs as little as ¥500. For a 100-episode series, that’s ¥50,000. Before AI, just the lead actor’s fee for that same series would have been more than the entire AI budget. The math is brutal, and it doesn’t care about your artistry.
But here’s the twist nobody’s talking about. These AI dramas aren’t better. They’re objectively worse — clunky shots, wooden expressions, zero cinematic craft. Yet they’re crushing the human industry because ‘worse’ content wins on volume, verticality, and platform economics. The algorithm doesn’t reward quality. It rewards cheap supply that feeds niche obsessions.
Walk into any short drama platform now and you’ll see the evidence. A homemade American high school AI romance. An AI horror series. A were-lion European gay drama, made by one person in their bedroom. These aren’t polished products. They’re what happens when someone says, ‘I want to watch my specific fantasy, and I’ll make it myself for pocket change.’
And that’s the deeper gut-punch. The human touch isn’t being outcompeted by perfection. The human touch is being outcompeted by cheap, unlimited, and deeply personal garbage. Garbage that speaks directly to one person’s XP — and they’ll pay a subscription for it. Garbage that creates private communities, bypassing traditional distribution entirely.
The actors aren’t losing to a better product. They’re losing to a million tiny mirrors reflecting individual desires back at themselves, each one costing less than a dinner bill.
But here’s where it gets dangerous for the platforms themselves. Right now, Red Fruit (the dominant short drama app) squeezes both creators and actors. The AI revolution started as a way for them to cut costs. But it’s also empowering a wave of solo creators who don’t need Red Fruit at all. They publish on Telegram, on Discord, on custom websites. They build audiences that cannot be scraped or algorithmically controlled.
The platform’s monopoly is both the cause of the current squeeze and the next bottleneck to break. The same AI that emptied Hengdian is now creating a distribution model that doesn’t need the gatekeepers. The actors are just the first casualties. The platforms themselves will be next.
So yes, fear the machine. But the more unsettling truth is that the machine isn’t coming to replace you with something better. It’s coming to replace you with something that costs nothing and tells exactly the story you secretly wanted to hear. And you’ll watch it. Because it’s cheap. Because it’s personal. Because it’s yours.
FAQ
Q: But can AI really replace the emotional depth of a real actor?
A: It doesn't have to. The economics of short dramas favor volume and niche targeting over emotional depth. AI doesn't need to be better—it just needs to be cheap enough to let a thousand micro-audiences get exactly what they want, even if that 'want' is objectively mediocre.
Q: What should short drama actors do now?
A: Two paths: either move up into premium, high-touch storytelling that AI cannot reproduce (live performances, complex improvisation), or lean into AI as a co-creator. The interim will be brutal, but actors who treat AI as a tool rather than an enemy—and build direct audience relationships—have a fighting chance.
Q: Isn't this just a fad? People will get tired of AI-generated content eventually.
A: That's wishful thinking. The cost advantage is permanent and compounding. Every month AI tools get cheaper and slightly better. The real shift is not in quality—it's in the economics of supply. Once a creator can serve a micro-niche profitably, they never go back. The fad is believing 'good enough' can't beat 'good'.