If you watched the NBA Finals on Wednesday night, you may have seen a surreal 30 seconds of AI-generated madness: an old man in a cowboy hat carrying a chihuahua, a swimmer in a pool of eggs, and an alien shotgunning a beer. It wasn’t a fever dream, it was a real ad from betting platform Kalshi. Cost: $2,000. Production team: one person. Tools: Google Veo 3, Gemini, CapCut.

Welcome to the era of “AI SlopComs” (aka: Generative AI Video Communications).

PJ Accetturo, who describes himself as an “AI filmmaker,” made the ad by generating 300–400 Veo clips and stitching together 15 usable ones. His creative process was algorithmically assisted: Gemini wrote prompts, Veo rendered nonsense, and CapCut assembled it into a coherent incoherence. If you can’t tell the difference between this and a “real” ad, is there a difference?

From a production standpoint, this is a 95% cost reduction over traditional workflows. From a narrative standpoint, it’s something else entirely. These AI videos mimic content language that’s been years in the making. The grammar is familiar: short 4–8 second clips, loosely connected, reliant on spectacle, surrealism, and the viewer’s willingness to make meaning.

We’ve seen grammatical shifts in video before. In the 1980s, music videos abandoned the tripod for handheld chaos. In 2008, mobile-first design gave rise to 9:16 vertical video. Vine condensed storytelling into 6-second loops. TikTok refined it. AI is now remixing it.

These ads don’t try to persuade. They demand attention. They’re optimized for scroll-stopping weirdness, not brand lift. And as production costs drop to near-zero, slop will scale.

This isn’t the future of storytelling. It’s the beginning of a strategic wave we can call “Use AI to make outrageous, attention-grabbing, WTF was that?” videos. It raises the same uncomfortable truth I wrote earlier this year: If You Can’t Tell the Difference, There Is No Difference.

More By This Author: