Why 'Niu Lai' Is Worse Than AI (And Why That's The Point)

There is a Chinese animated film this summer that looks, to the untrained eye, like it was assembled in a weekend by someone's first 3D-modeling tutorial. The cow has a face that defies easy description. The background grass appears to be clumps of green mosaic. Characters slide more than they walk. And yet, against every rule of the box office, it became one of the year's unlikely hits. The rough look is the reason, not a flaw the film overcame.

The film is called Niu Lai. The strangest part of its moment is a sentence that keeps turning up in comment sections, translated and retranslated across languages: it is worse than AI.

On a technical level, that sentence is almost certainly true.

The machine could do better in seconds

It is 2026. A person with no training can open a generation tool and produce, within minutes, animation smoother than anything in this film. The characters move naturally. The light falls correctly. There are no visible seams, no stiff transitions, no cattle that look like they were built from leftover geometry. By any normal measure of craft, an AI-assisted pipeline would beat what two people made over several years of evenings and weekends.

So when viewers call the film "worse than AI," they are not wrong about the pixels. They are pointing at something real. In a contest of pure technical polish, the amateurs lost.

That loss is where things get interesting.

Imperfection as a fingerprint

For most of the last decade we treated flawlessness as the goal. Smoother renders, cleaner cuts, more convincing faces. Then generative tools arrived and made flawlessness cheap. Practically free. The market flooded with content that was technically competent and emotionally identical. Polished, weightless, interchangeable. The slang for it is blunt: slop.

Into that flood walks a film that is obviously, embarrassingly handmade. Every awkward frame is a fingerprint. The stiffness is not a style borrowed from a trend. It is the literal limit of what two people with secondhand computers could produce. You can feel the effort in the mistakes.

That is the inversion at the core of the Niu Lai moment. When machines can make anything perfect, perfection stops being rare. What gets rare is the visible trace of a human who tried, fell short of elegance, and shipped the result anyway. The film's badness is not a defect to hide. It is the proof that a person was here.

The audience defended the original against its own upgrade

The clearest evidence came from the fans. After the film went viral, users ran its crudest scenes through AI tools and produced remastered versions: smooth, clean, professional-looking. These upgraded clips were technically superior in every way.

The audience rejected them.

Again and again, viewers said the polished versions felt empty, that something essential had been erased. The rough original, they argued, carried a warmth the algorithm could not reproduce. It is a remarkable thing. People were offered a better-looking film for free and chose the worse one on purpose, because the worse one felt alive.

That choice says less about the movie than about the moment we are in. We are hungry for work that announces, in its very flaws, that a person made it.

A mother, a son, and several years of stubbornness

The story behind the film sharpens the point. It was made by a mother and her adult son, neither with any background in animation. They taught themselves. They wrote it, modeled it, edited it, and voiced it. They used equipment they could afford, much of it used. The process took several years.

None of this makes the result good by industry standards. But it reframes the awkwardness. The stiffness is no longer just "bad animation." It is the residue of two people learning a craft from zero, refusing to quit, and finishing something enormous by hand. Once you know that, the frozen frames stop reading as lazy and start reading as stubborn.

Cinemas, caught without promotional artwork, reportedly handed markers to their staff and let them draw posters by hand: crude cows, handwritten slogans, a straight-faced "no refunds." Those posters became attractions in their own right. The whole episode turned into a nationwide, low-budget improvisation, and the audience loved it precisely because it was unscripted and human.

An honest caveat

It would be dishonest to pretend this is a great film, or even a good one by normal measures. Plenty of viewers find it unwatchable, and some argue the "handmade hero" narrative overlooks how uneven the result is. There is also a fair reading that the film's rise was less a rebellion and more a giant inside joke. Millions of people bought a ticket mostly to see how bad it really was, then bonded over the shared spectacle.

And the "anti-AI" label is something the audience projected onto the film more than anything the filmmakers claimed. The movie was not made as a statement about artificial intelligence. It was made because two people wanted to finish a story. The meaning was supplied afterward, by a public starved for anything that did not feel generated.

That distinction matters. The value is not in the film's quality. It is in what its reception reveals about us. In a year when machines can fake perfection on demand, we are moved, almost against ourselves, by the unmistakable mark of a human hand, even (or especially) when that hand is clumsy.

The point

So yes, Niu Lai is worse than AI. A model could render a prettier cow in the time it takes to read this sentence. But the model can't show you the several years a mother and son spent refusing to let an impossible idea die. There is no algorithm for the slightly unhinged pride of a theater employee's felt-tip cow, or for the relief of a room full of strangers laughing at something gloriously, defiantly unfinished.

We are building a world where almost anything can be generated and polished on demand. Quality stops being scarce. What gets scarce is authenticity, the evidence that a real person was here, tried, and left a mark.

Niu Lai is rough and ridiculous, and most people won't watch it twice. But as a mirror held up to 2026, it cuts sharper than a lot of far more competent work, because it proves something a polished film cannot. Someone actually made this.