How One Vocational Graduate Used AI to Reach 100 Million Views

The Unexpected Breakout

In May, a short film called Zombie Scavenger began circulating on Chinese social platforms with unusual velocity. Within weeks, it had crossed 100 million cumulative views. The director was not an alumni of the Beijing Film Academy or a studio apprentice. Liu Ziyu graduated from a vocational college, worked in real estate in Yuxi - a mid-tier city in Yunnan province - and had no prior credits in traditional production.

What he did have was access to generative video models and a narrative instinct honed by watching Love, Death & Robots and other anthology sci-fi. The result was a visually ambitious piece that domestic audiences compared favorably to Netflix's celebrated series. More significantly, it caught the attention of a Hollywood producer, who extended an offer to Liu. At DailyTechWire, we've tracked the emergence of AI-native creators across Southeast Asia and East Asia over the past eighteen months, but few have achieved Liu's combination of scale and industry recognition this quickly.

The story raises a set of interlocking questions: Are we witnessing a genuine democratization of filmmaking, or simply a new layer of gatekeeping where prompt engineering and compute access replace camera crews and editing suites? And if the barrier to entry has dropped, what happens to the established pipeline that has long controlled who gets to direct?

Tools That Lower the Floor

Generative video platforms - ranging from open-weight models fine-tuned on local hardware to commercial APIs offering scene synthesis and character animation - have matured rapidly in the past year. Liu's workflow, as he described in follow-up interviews on Chinese social media, involved scripting scenes in natural language, iterating on visual outputs, and stitching sequences together with conventional editing software. The process still required creative judgment, but it bypassed the need for physical sets, actors, or lighting rigs.

This technical shift mirrors what happened in music production a decade ago, when digital audio workstations and sample libraries allowed bedroom producers to compete with studio professionals. The difference is speed: video generation has compressed a similar trajectory into roughly two years. For aspiring filmmakers in regions where production infrastructure is sparse or prohibitively expensive, the implications are immediate.

In Indonesia, Vietnam, and the Philippines, we've observed small cohorts experimenting with similar hybrid workflows - part AI synthesis, part manual post-production. Few have achieved Liu's reach, but the pattern is consistent: creators who would have faced insurmountable capital requirements now have a path to prototype and distribute work at minimal cost.

The Hollywood Knock

The Hollywood producer's offer to Liu is noteworthy less for its novelty - studios have long scouted talent from viral hits - than for its timing. It signals that North American gatekeepers are beginning to treat AI-native work as a legitimate portfolio rather than a curiosity. Whether that translates into sustained career opportunities or one-off contracts remains to be seen.

Traditional film schools and apprenticeship systems have served as credentialing mechanisms, filtering candidates through years of training and networking. Liu's trajectory bypasses that entirely. His credibility comes from audience metrics and the technical competence demonstrated in the finished piece. If this becomes a repeatable pattern, it threatens to disintermediate not just production companies but also the educational institutions that have long served as talent pipelines.

At the same time, the offer raises questions about exploitation. Studios may view AI-literate directors as cost-efficient hires who can deliver high visual output with smaller crews and budgets. That could open doors for individuals like Liu, but it could also erode labor standards and push downward pressure on rates for conventional directors and crew members.

The Economics of Attention

Zombie Scavenger succeeded in part because it arrived at a moment when Chinese audiences were primed for domestically produced sci-fi that felt visually competitive with Western imports. The film's aesthetic - gritty, post-apocalyptic, kinetic - resonated with viewers who had grown accustomed to the visual language of Love, Death & Robots but wanted stories rooted in their own cultural context.

Liu's background in real estate, a sector known for its relentless sales cycles and performance pressure, may have given him an intuitive grasp of what would capture attention in a crowded feed. The film's pacing, its hook in the opening seconds, and its shareable moments all suggest someone who understands how content propagates on platforms optimized for short-form engagement.

This is not incidental. The virality of AI-generated or AI-assisted work often depends less on traditional narrative craft and more on platform literacy - knowing how to structure a piece so that it survives the first three seconds, generates comments, and gets pushed by recommendation algorithms. Liu's success is as much about understanding distribution as it is about mastering generation tools.

Who Else Is Waiting?

Liu's story has already inspired a wave of imitators. Forums and chat groups focused on AI filmmaking in China have seen a surge in activity, with users sharing prompt techniques, model recommendations, and distribution strategies. The narrative of the vocational graduate who leapfrogs the system is powerful, and it's being amplified by platform algorithms that reward underdog stories.

But replication is not guaranteed. Liu benefited from timing, a specific cultural moment, and a story that aligned with audience appetite. The next cohort of AI-native filmmakers will face a more saturated landscape, where novelty has worn off and expectations have risen. The tools may be accessible, but the attention economy remains zero-sum.

There is also the question of sustainability. A single viral hit does not constitute a career. Liu's next project will be scrutinized not as a proof-of-concept but as evidence of whether he can repeat the formula or evolve beyond it. Hollywood's interest may evaporate if follow-up work fails to deliver similar metrics or if the novelty of AI-generated content fades.

What Changes, What Doesn't

The arrival of accessible generative tools does not erase the fundamentals of storytelling - character, tension, pacing, emotional resonance. Liu's film succeeded because it delivered on those fronts, not merely because it was made with AI. The technology provided leverage, but it did not substitute for narrative instinct.

What does change is the composition of who gets to attempt the work. For decades, filmmaking has been a high-capital, high-credentialing field. The barriers were not just creative but structural: access to equipment, access to collaborators, access to distribution. Generative models and social platforms have lowered some of those barriers, though not eliminated them. Compute costs, platform algorithm dynamics, and the need for post-production skills still favor those with resources and technical fluency.

The broader question is whether this shift will diversify the range of stories being told or simply replicate existing power structures under a new technical surface. If AI filmmaking becomes dominated by a small number of well-capitalized creators or platforms that control distribution, the democratization narrative will prove hollow. If it genuinely enables voices like Liu's - people outside traditional pipelines, working in second- and third-tier cities, telling stories that would not have been greenlit by conventional studios - then the transformation is real.

Forward View

Liu Ziyu's trajectory from Yuxi to Hollywood pitch meetings is a data point, not yet a trend. But it's a data point that studios, film schools, and aspiring directors across Asia are watching closely. The tools that enabled Zombie Scavenger are now available to millions. Whether that leads to a Cambrian explosion of new voices or a glut of low-quality content competing for the same shrinking attention span will depend on how the economics of creation and distribution evolve over the next twelve to eighteen months.

For now, the lesson is clear: the path into the industry is no longer singular. Traditional routes still exist, but they are no longer the only routes. And for those willing to learn the grammar of generative tools and the dynamics of platform distribution, the gap between ambition and execution has narrowed considerably.