In 23 days, one creator without formal filmmaking training made 1 million yuan—roughly $138,000—from short-form AI-generated video content. Director Guo Fan, whose credits include *The Wandering Earth*, publicly praised the achievement. The case has become a rallying point for a fundamental shift underway across creative industries: humans are no longer the pixel-pushers. They are becoming creative directors overseeing systems that do the technical work.
The transformation is visible in the latest version of Pixcake, a professional photo retouching tool popular among Chinese commercial photographers. Version 9.0, released this week by the Beijing-based team behind the software, represents what the company calls a "role inversion." Where earlier versions gave humans more filters to click and sliders to adjust, Pixcake 9.0 lets the AI handle technical execution—the labor-intensive work of skin smoothing, color grading, background replacement—while the human evaluates results and makes creative decisions. The retoucher becomes a manager reviewing AI output, not a technician executing commands pixel by pixel.
"The retoucher is shifting from 'operator' to 'manager,' from 'technical executor' to 'creative decision-maker,'" the Pixcake team explained in their announcement. This mirrors what happened in architecture when CAD software eliminated the need for hand-drafted blueprints: junior drafters didn't disappear, but their role transformed entirely.
The creator who earned that million-yuan milestone exemplifies the same inversion at a different scale. Without a film school background or a professional crew, they assembled finished short videos using AI generation tools, then focused their energy on story concept, pacing decisions, and audience research—the distinctly human work of knowing what a viewer wants to feel. The AI handled video generation, scene transitions, and visual effects. The creator directed.
This pattern is appearing across creative sectors. Fashion photographers who once spent hours on complex backdrop compositing now describe their workflow as "curating AI outputs." Graphic designers at mid-size agencies report that clients increasingly ask them to "just use AI" for technical execution so the humans can focus on strategy. The economic logic is stark: a single creative director can oversee multiple AI production pipelines that would have once required a team.
The implications for creative professionals are neither uniformly dark nor uniformly bright. Practitioners who adapt quickly—learning to evaluate AI output critically, to write effective prompts, to maintain a coherent creative vision across AI-assisted production—may find their leverage increasing. Those whose skills are concentrated in technical execution rather than creative judgment face real pressure.
Pixcake's own user community reveals this bifurcation. Power users celebrate the productivity gains: what once took three hours of detailed retouching now takes thirty minutes of review and refinement. But a quieter cohort asks what happens when the skill of detailed retouching—built over years of practice—becomes irrelevant to the market. The company's answer is characteristically pragmatic: learn to direct, not to execute.
Guo Fan's endorsement matters beyond the publicity. As someone who controls major production resources, his implicit message—that AI-generated content can meet professional standards—signals that the technology is crossing from hobbyist novelty into legitimate production tool. When a director of his stature says a short video creator's AI workflow "deserves attention," the industry takes notice.
The 23-day million-yuan number will fade into context. What persists is the structural insight it illustrates: the bottleneck in creative production is shifting from execution to conception, from knowing how to make something to knowing what to make and why.