Can AI tell if your script will make a hit film?
Original reporting by The Verge
The AI startup Quilty burst onto the scene this year with a bold promise: its tool could accurately predict a film's box office success simply by analyzing its script. Pitched as a democratizing force for the entertainment industry, offering aspiring creatives invaluable foresight, the technology quickly faced an acid test. Early experiments revealed a glaring flaw, as Quilty notoriously predicted an eventual box office flop, *Christy*, would outperform the Oscar-winning blockbuster *Sinners*. This immediate misstep cast a shadow over its ambitious claims.
A new approach
Founded by film producers Simon Horsman and Daniel Wood, Quilty aims to provide detailed reports and a score from 0-100 on a script’s narrative quality, commercial viability, and audience resonance. Crucially, its method is not built on a singular, bespoke AI model. Instead, Quilty stitches together an array of widely available AI tools—like Gemini for structure, DeepSeek for financial modeling, and a combination of Claude and ChatGPT for character analysis. While this modular approach allows for agile updates, it also raises fundamental questions about its capacity to truly comprehend the intangible elements that define a hit film. The company maintains its goal is to "keep humans in the loop," but its inability to foresee crucial external factors or grasp nuanced cultural shifts leaves its predictive power in doubt, underscoring the deep chasm between sophisticated pattern recognition and genuine artistic insight.
Quilty's ambitious premise—to quantify the elusive magic of a hit film script—ultimately runs aground on the complex, often unpredictable nature of human taste and cultural zeitgeist. While its founders laud its modularity and commitment to "human-in-the-loop" development, the tool’s inconsistent predictions, like favoring the box office flop *Christy* over the Oscar-winning *Sinners*, underscore a fundamental challenge. Current AI, built on pattern recognition and sophisticated mimicry, demonstrably lacks the nuanced comprehension required to truly understand artistic merit or foresee the myriad external variables that impact a film’s success. The $50 price point for a composite analysis of readily available models suggests a convenient aggregation rather than a breakthrough in genuinely predictive capability.
AI's Creative Frontier
This episode with Quilty is a microcosm of a larger truth emerging in the AI landscape. While generative AI excels at optimizing certain processes and identifying trends in structured data, its limitations become starkly apparent when grappling with subjective creative output. The promise of "democratizing" the film industry through such tools, by offering a definitive score or a glimpse into the future, remains a distant prospect. This highlights that genuine artistic insight and the capacity to predict audience reception still firmly reside with human intuition, experience, and the inexplicable spark of creativity. For AI to truly become an integral, reliable partner in creative development, it must evolve beyond sentiment analysis and statistical pattern matching, making a leap into something closer to genuine comprehension—a significant hurdle for the technology in its current form. Until then, tools like Quilty serve as a reminder that the most compelling stories, and their potential for success, defy purely algorithmic definition, continuing to emphasize the irreplaceable role of human judgment.