Filmmaking Beyond the ‘Slop’ of Generative AI

For the better part of two years, the dominant conversation around AI-generated imagery has been summed up in a single derogatory word: slop. It has become a catch-all term used to describe uncanny, mindless, visually assaulting images and videos generated by LLMs from user prompts and published promptly online and on social media.  But is it fair to call all AI-generated content slop?  Google Labs is trying to rewrite that narrative with its Flow Sessions. At New York’s Metrograph theater on Wednesday, Google screened several short films created by the fourth cohort of its Google Flow Sessions artist residency program.  I got an early look at the films created by each of the 11 resident artists. Far from the mindless output that the term “AI slop” implies, the debuting works are deeply and intensely emotional, personal, tactile and imaginative. As someone who spent years taking film classes and studying avant-garde cinema, these projects feel less like traditional shorts and more like cinematic memories or “essay films” that incorporate the use of generative AI tools during production, not as a shortcut to the scrutinous creation and editing process, but as an assistant to create visions that previously required prohibitive budgets and large-scale crews.  At the center of these short films is Google Flow, a generative AI media creation tool designed to serve as a collaborative canvas that brings an artist’s vision to life. Through Flow Sessions, Google organizes a six-week residency program that brings together artists from a range of backgrounds, including fine art, commercial agencies and beyond, to develop passion projects. Brianna Doyle, head of community at Google Labs, told me that Flow Sessions’ mission is to close the gap between the developers who write the code behind these tools and the artists who actually use them.  The aim isn’t to build AI for artists, but to build AI with them, so the tools become a conduit for expression rather than a replacement for human creativity, storytelling and process.  “AI has opened the doors for them as far as creativity goes,” Doyle said. “It hasn’t been stifling at all. It’s actually reframed what they are able to do.” Lowering the floor and raising the ceiling  The dominant fear among traditionalists is that the use of AI dilutes authenticity, reducing the sanctity of art to machine-made approximations. But for the artists in this cohort, the technology has worked as a catalyst for a new – yet still nuanced – kind of storytelling. The motto, as Doyle says, is to “lower the floor and raise the ceiling.” It’s about reducing the barrier to entry for newcomers to filmmaking while expanding what an independent artist can make and refine. That shift shows up in the way the tool itself has evolved over time. Flow began as a video-focused system, but artists reported spending most of their time wrestling with images and storyboards rather than footage. In response, Flow incorporated image generation and an organizational interface — folders, labels and structure – that helps ideas move from concept to production without losing momentum, according to Doyle.  For the resident artists, like Ikenna Mokwe, founder of OnGen Studio, that efficiency changes what’s possible. After 15 years in traditional media, he has developed a new workflow in which LLMs and video generation help bridge the gap between an abstract pitch and his vision. His process is built as a loop: he uses Gemini via voice commands to transform late-night brain dumps into structured scripts, cutting out the physical bottleneck of typing, and then relies on custom micro-tools within the flow to automate repetitive camera movements.  Still, Mokwe is clear about what matters most. “Storytelling is at the heart of everything,” he said.  To Mokwe, AI doesn’t write the story, but it provides a faster path from conception to completion.   “I’ve had absolutely zero hesitation,” he sai