I Caught Myself Watching a Scripted Kitchen Fight, Twice

Last night, lying in bed scrolling my phone, I hit a video: a couple in their kitchen splitting a carton of eggs, the husband counting out that his share was one short, flipping out on the spot. Camera shoved in his face, comments piling on — "what's wrong with this guy." I watched it twice, and the more I watched, the more it didn't sit right. Two people supposedly mid-fight, and the camera stayed rock steady, like it was bolted to a tripod — nobody touched it. The lighting was even, none of the mess a real kitchen fight would have on anyone's face. This was almost certainly scripted.

This kind of content — the kind that makes you want to comment something angry the moment you see it — has a name: rage bait. Oxford's word of the year for 2025. The formula isn't complicated: stage a conflict vivid enough that people get angry before they can tell it's fake, and anger gets you clicks, comments, shares. Some accounts go further — putting the "close" button right next to the "like" button, so a slip of the thumb becomes an engagement.

That gut-drop I felt watching the egg-splitting video wasn't about the content — it was about catching myself in the mirror. I write for a living too. Every day I'm asking myself the same question: is this line sharp enough to get people talking, will this phrasing make someone want to push back. Watching someone else run the script this openly was like watching my own process from the outside.

Conflict works, and not just because people like watching a fight. There's a deeper mechanic underneath it: in a society of strangers, whoever airs the grievance first, wins. It's the same logic as the prisoner's dilemma — two suspects held apart, whoever talks first gets the lighter sentence, whoever stays quiet eats the loss, neither trusts the other to hold out, so both rush to talk. When people lived in villages or work-unit compounds, neighbors and coworkers had to deal with each other for decades — whoever broke face first lost standing, so grievances mostly got swallowed. Once everyone moved into anonymous apartment blocks where nobody knows anybody, that restraint disappeared. Now, whoever cries foul first gets to look like the wronged party.

Content creators are the group that pushed this logic to its limit: they don't need to wait for a real conflict to happen. They can script one directly, and it works just as well as the real thing — faster and more reliably, too. The couple splitting eggs probably weren't a real couple. But the thousands of angry comments under that video were real.

This path isn't as smooth as it looks. Two things are closing in on it.

One is the platforms themselves. People who study public opinion trends have found that recommendation algorithms aren't as crude as "the angrier it is, the more it gets pushed" anymore. They run multi-objective models now — ones built to specifically identify content tagged as genuinely abusive or toxic, and actively downrank it. Which means pure shouting-match content has a narrowing lane for growth; what keeps pulling sustained traffic is content that sits in the gap between "looks like conflict" and "crosses the line" — leaning on pure hostility backfires.

The other is the law. Just this past June, there was a case of a staged conflict that blew up in the creator's face: a content creator used AI to face-swap, cut footage out of context, and fabricate dialogue, turning a normal business interview into a fake clip claiming a beverage company's founder "refuses to drink his own product." Within days, the company's shares on both the A-share and Hong Kong markets dropped, wiping out over 7 billion yuan in combined market value within a week. Not long after, police in Shenzhen announced the person who edited the clip had been criminally detained. Whatever traffic a staged script can pull in is starting to look like a much worse trade than the criminal liability it can pull in.

I've written for a living long enough now that I cross this line every day — whether a sentence should lay out the reasoning clearly, or get people fired up first. I understand the prisoner's dilemma logic; whoever airs it first takes the lead, and when I write a "judgment," am I not doing some version of the same thing — trying to get there first? But scripting a conflict that never happened, and stating a real judgment as directly as I can, are two different things — and I know which one I'm doing.

Next time you hit a video that's sitting right on that edge where you want to type something angry — do you click in, or do you scroll past?