'Digging the grave of my profession': the Hollywood creatives training AI to do their jobs
Award-winning Hollywood writers, directors, and producers are taking gig work training AI models for companies like Anthropic and OpenAI, earning $12–$200/hour to offset declining industry income The trend reflects a severe jobs slump in Hollywood, with LA shoot days falling 48% between 2021–2025 and US motion picture jobs declining 28% from 2022 to 2026 Training tasks include teaching AI to devise production schedules, create pitch decks, transcribe video footage with overlapping audio, and eva
Analysis
TL;DR
- Award-winning Hollywood writers, directors, and producers are taking gig work training AI models for companies like Anthropic and OpenAI, earning $12–$200/hour to offset declining industry income
- The trend reflects a severe jobs slump in Hollywood, with LA shoot days falling 48% between 2021–2025 and US motion picture jobs declining 28% from 2022 to 2026
- Training tasks include teaching AI to devise production schedules, create pitch decks, transcribe video footage with overlapping audio, and evaluate AI-generated creative content
- Prominent AI training agencies like Mercor, Micro1, and Handshake are actively recruiting experienced creatives with 5+ years of credited project experience for up to $85/hour
- Creators express mixed feelings—ranging from fatalism and guilt to cautious optimism—about helping build the technology that could replace their professions
Why It Matters
This article highlights a critical inflection point where AI adoption is not just threatening creative industries but actively recruiting their most skilled practitioners to build the very systems that could displace them. For AI practitioners and industry observers, it underscores the accelerating convergence of generative AI with knowledge-work domains and the complex ethical and economic dynamics emerging as talent shortages in AI training drive companies to poach from the industries they are disrupting.
Technical Details
- AI training tasks involve creating evaluation datasets and simulation scenarios for production-management workflows, including budget reconciliation, scheduling adjustments, vendor/crew coordination, and permit identification
- One documented task required an AI to generate detailed two-day shoot schedules incorporating filming permits, location hazards, daylight conditions, personnel lists per shot, and cast needs such as child protection compliance
- Another task involved teaching AI to transcribe video with overlapping audio (e.g., crowd noise and music at a Little League game), requiring creation of visual speaker guides and accent characterization across English, Scottish, African American, and Asian dialects
- Training agencies like Micro1 are designing tasks that simulate realistic production-management scenarios, requiring creators with demonstrated mastery in film, television, digital, or live event production
- Netflix reported using AI in 300 of its 1,000 titles in 2026, signaling large-scale production integration of AI tools
Industry Insight
- The "shovel seller" paradox—where displaced workers train the AI that displaces them—will likely intensify as AI companies face talent shortages and turn to domain experts for high-quality training data, creating ethical dilemmas for both workers and companies
- The Hollywood jobs decline (48% drop in LA shoot days, 28% national employment drop) demonstrates how AI disruption compounds existing industry headwinds like pandemic aftershocks, strikes, and streaming investment cuts, suggesting similar patterns will emerge in other creative and knowledge-work sectors
- While some creators remain optimistic that AI cannot yet produce award-winning work, the rapid integration by major studios (Netflix, Ron Howard's AI-enabled documentary) signals that the technology is moving from novelty to production necessity, making adaptation and upskilling essential for creative professionals
Disclaimer: The above content is generated by AI and is for reference only.