36 Hours, 200 Million Views: AI Video’s Moat Isn’t in the Tools

An account appeared 40 days ago and got about 200 million views across platforms in 36 hours. The work is called "Englishman in New York," and the character is a British gentleman in New York.

Linus Ekenstam mentions in his tweet that @antrofrog_, labeled a "zero-skill artist," recently emerged on social platforms over the past 40 days. This video got about 200 million views across platforms within 36 hours. The production tools are no mystery: character design with Midjourney 8.2, animation with Seedance 2.5, completed on Krea AI.

The short film’s structure is clear. Each segment has a hook: it builds suspense first, then releases it, and the next segment layers on, while never straying from the premise of “a British gentleman in New York.” Linus speculates that the music is very likely driving the story. The whole short ends on a moral line: a gentleman doesn’t have to stand in line. If you want to cut in line, you don’t have to be a jerk—do something good, be a good person.

This ending isn’t complicated, but it gives all the earlier small segments a sense of closure. A common problem with short videos is that they have only satisfying moments but no landing point; this film makes its landing point very light.

More importantly, it isn’t a one-off hit on a single platform. Linus says this video tends to take off wherever it’s posted: it blows up when Antrofrog posts it, and it also blows up when others download and re-upload it. Within 5 hours, Antrofrog posted a second video with the same character on Instagram; the second has 62 million views so far and is still rising. The Instagram account now has 325,000 followers.

Linus’s judgment is that the tool barrier has fallen to a historic low, but the real skills of storytelling and delivery remain the biggest moat. You can call it taste, because luck cannot explain the spread of two consecutive videos.

In the comments, someone put it more directly. A netizen said that “zero skill” means no traditional drawing and visual production skills, but directing ability is 100% present. This distinction matters. AI has lowered the barrier to drawing and animation generation, but storyboarding, pacing, character consistency, music selection, and platform adaptation are still in human hands.

Someone asked whether Antrofrog could reveal the workflow. Antrofrog replied that it’s the same as everyone’s workflow, nothing special. This answer sounds like a brush-off, but it also shows that the tool combination itself isn’t a secret. Midjourney for characters, Seedance for animation, Krea to connect them—many people can reproduce this process. What’s hard is why each segment stops where it does, why the music enters, and why the next hook can catch.

Linus also posted two screenshots in response to the discussion.

There’s another question in the comments worth copying: Do you choose the music first and then make the visuals? This question is closer to the core of virality than asking about model parameters. The rhythm of short videos is often set by sound; visuals follow the music, and edit points become much more natural. Linus didn’t give a definitive answer; he only mentioned that on Instagram the music comes from Reels’ searchable music library.

So the easiest way to misread this case is to treat it as evidence that “zero skill can also go viral.” A more accurate view is that tools have filled in part of traditional visual skill, while narrative and directing ability have been amplified. Antrofrog spent 40 days building a character, continuously testing, and distributing across platforms, ultimately getting 200 million views and 325,000 followers. After tools become equalized, the gap is in taste and execution.

Source: Linus Ekenstam tweet https://x.com/LinusEkenstam/status/2116850561832149323 , Antrofrog original post https://x.com/antrofrog_/status/2116118341610004764 .