Entertainment workers are increasingly turning to AI training gigs with companies like Mercor as a primary income source. The work, which involves labeling data and refining AI outputs, has become the modern equivalent of waiting tables for struggling writers in Hollywood.
According to a first-hand account from screenwriter Ruth Fowler in Wired, AI training work offers flexible, remote employment for creative professionals facing industry instability. The gig involves tasks such as evaluating AI-generated content and providing feedback to improve language models.
Companies like Mercor connect workers with AI training projects, compensating them for their expertise in narrative, dialogue, and storytelling. The work appeals to entertainment professionals with gaps between projects or facing reduced opportunities.
While these gigs provide immediate income, they raise questions about fair compensation and the broader implications of using creative workers to train the AI systems that may eventually compete with their original work. The trend reflects how economic pressures are reshaping career paths in entertainment, with traditional stepping-stone jobs being replaced by AI-focused contract work.
As the entertainment industry grapples with strikes and uncertain employment, such gig work has become a financial necessity rather than a temporary measure for many writers.
The Pentagon has launched customized versions of OpenAI's ChatGPT and xAI's Grok on its GenAI.mil platform, providing 3 million military and civilian personnel with AI tools designed for defense operations.
Instagram is restricting the visibility of artificial intelligence accounts that don't disclose their AI nature, responding to growing user frustration with AI influencers.
A new initiative encourages workers to avoid AI tools one day per week. The movement has sparked debate on tech community forums with 174 upvotes on Hacker News.
Instagram is replacing its "AI creator" tag with a clearer "AI-generated profile" label after acknowledging that users frequently cannot distinguish artificial accounts from real people.