The rise of AI-generated films has been a controversial topic in the industry, with debates raging over creativity, authenticity, and whether machine-generated storytelling can ever truly match human ingenuity. Nobody Believes Cassandra, the latest AI-assisted sci-fi thriller directed by James G. Maynard, ambitiously pushes the boundaries of AI filmmaking. While it stands as a notable advancement in AI cinema, the film itself is a mixed bag—brimming with fascinating concepts yet falling short in execution.
At its core, Nobody Believes Cassandra tells the story of Dr. Cassandra Selene, an AI researcher who develops HELEN, an artificial intelligence designed to decode historical mysteries. However, as HELEN begins predicting the future, including a catastrophic earthquake in Mexico, Cassandra is met with skepticism and resistance from governments, corporations, and the media. The film positions itself as a high-concept thriller exploring AI’s potential and the societal consequences of ignoring forewarnings.
One of the film’s strongest aspects is its thought-provoking premise. AI predicting disasters and being dismissed by those in power is a timely and compelling theme. The film effectively raises questions about technological advancements, misinformation, and the fine line between scientific progress and fearmongering.
Moreover, Cassandra's struggle is deeply resonant—whether AI-generated or not, the script presents her as a compelling protagonist whose frustrations with bureaucracy and corporate greed feel authentic. The film’s narrative structure is also commendable, mirroring elements of ancient mythology while keeping the stakes contemporary and urgent.
Visually, Nobody Believes Cassandra showcases incremental improvements in AI-assisted filmmaking. The character designs are more refined than previous AI-generated films, though some still fall into the uncanny valley. One standout is Apollo, a character who looks like a mix of Ethan Embry and Cam Gigandet, which makes me curious about the dataset used to generate his likeness.
However, the film relies heavily on talking-head sequences, with minimal variation in cinematography. While AI can generate competent visuals, it struggles with dynamic composition and scene transitions. The latter often feel abrupt, reminiscent of Star Wars-style wipes, though they don't quite fit the film’s tone.
The dialogue is another mixed bag. At times, the interactions between Cassandra and HELEN are engaging, particularly in their philosophical discussions about time and probability. But elsewhere, conversations feel stilted, as if AI-generated text wasn’t refined enough in the editing process. The film’s humor occasionally lands, but there are also moments where the delivery lacks naturalism.
For all its ambitious ideas, Nobody Believes Cassandra still doesn’t fully function as a film in the traditional sense. The lack of visual variety and over-reliance on exposition makes it feel more like an elaborate AI experiment than a fully realized cinematic experience. While Cassandra’s journey is intriguing, the film fails to build emotional engagement beyond its surface-level conflicts.
Additionally, Maynard’s role in the production raises questions about authorship in AI filmmaking. Despite the heavy reliance on AI tools, he credits himself as a writer, director, and composer. Would “creator” be a more accurate title? These uncertainties highlight the ongoing challenges in defining artistic ownership in AI-assisted media.
Nobody Believes Cassandra is undeniably a step forward for AI-generated cinema, demonstrating the technology’s evolving capabilities. However, it also serves as a reminder that AI alone cannot create a compelling film without significant human refinement. The film contains intriguing ideas and moments of brilliance but ultimately lacks the cinematic craftsmanship needed to elevate it beyond technological curiosity.
While this is a step above the last AI film I watched, the art form (if you can call it that) still has a long way to go. This isn’t a great film, but it is an important one—a timestamp of where AI filmmaking stands in 2025. With further refinement, perhaps AI-assisted storytelling will someday craft a film that’s not just impressive for how it was made, but genuinely engaging in its own right.
Jessie Hobson