I looked inside an AI generated movie, and the best parts were all human

A short film made with Higgsfield, an AI video generation platform, has become a case study in what it genuinely means to make a movie with generative tools right now. The project follows three bumbling English lads who nurse pints in a grimy London pub while daydreaming about fame, groupies, and yacht parties - a premise that sits somewhere between Edgar Wright and Guy Ritchie in tone. The Verge spent time with the production to get an honest look at what AI contributed and what it could not.

The film, sourced from a Black List script, leans on Higgsfield's video generation capabilities for its hyperkinetic action sequences - smash cuts, chase scenes, and the kind of kinetic visual energy that would traditionally require significant crew and budget. Higgsfield is one of a growing number of platforms positioning generative video as a practical production tool rather than a novelty, and this project was intended to stress-test that proposition against a real narrative script.

What The Verge found, however, is that the most compelling moments in the film trace back to human decisions - the specificity of the characters, the timing of the comedy, the texture of the writing. AI-generated footage can approximate style and action at speed, but it struggles to carry the idiosyncratic weight of performance and character that makes a scene actually land. The technology functions more as a capable production accelerant than as a creative force on its own.

That distinction matters as studios, independent filmmakers, and platforms debate how generative video fits into real workflows. The film is a useful data point precisely because it does not pretend the technology is seamless - it shows both what becomes easier and what remains stubbornly dependent on human judgment. For anyone tracking the practical state of AI video generation, this production offers a grounded, if modest, proof of concept.