‘Dreams of Violets’ Review: What Does a Film Made Entirely with AI Look Like? Ash Koosha’s Iranian Protest Drama Is Dramatically Numbing, but It’s Still a Startling Portent of the Future

The release of *Dreams of Violets* isn’t just another curiosity in the expanding frontier of artificial intelligence and media—it’s a glimpse into a future where the line between machine-generated art and human storytelling grows increasingly porous. Ash Koosha’s AI-driven protest drama may not deliver a cinematic experience that rivals traditional filmmaking in emotional depth or nuance, but its existence forces a reckoning with what it means to create at scale, without the constraints of budgets, locations, or even human labor. The film’s reliance on generative AI to populate streets with soldiers and capture the chaos of Tehran in wartime underscores a broader shift: the erosion of barriers that once separated the possible from the speculative in visual storytelling. What makes this project particularly noteworthy is its timing. In an era where disinformation and synthetic media already blur public perception, *Dreams of Violets* arrives at a cultural inflection point. Audiences are growing accustomed to deepfakes, AI-generated imagery, and algorithmically curated content, but a full-length film constructed entirely through machine learning pushes the technology further—into the realm of narrative and atmosphere. The film’s static, almost numbing execution may betray the limitations of current AI tools, but it also highlights how quickly these systems are evolving. Already, AI can generate photorealistic environments and populate them with plausible characters, raising questions about the future of filmmaking: Will directors soon be able to "direct" AI actors and sets with the same precision as traditional crews? Could this democratize filmmaking—or further concentrate creative power in the hands of those who control the underlying models? Yet the most pressing questions linger beyond aesthetics. How will audiences distinguish between AI-generated content and authentic human expression when the technology matures? Will regulators step in to demand disclosure standards for AI-crafted media? And perhaps most critically, as synthetic realism becomes indistinguishable from reality, what does it mean for the role of film as a medium of truth, emotion, and human experience? Koosha’s work may be a clunky first step, but its implications are anything but. The film doesn’t just ask what AI can make—it asks what society will do when anyone, anywhere, can generate a world on screen without ever setting foot on a set.