Patrick Ciriello has a master's degree in information management. He spent decades designing software systems for banks and universities. At 60, after a year of unemployment that left his family sleeping in a Toyota Highlander in Walmart parking lots, he answered a cryptic LinkedIn message for a 'content writer' job. It turned out to be AI training work. He now reviews AI responses for errors at $21 an hour, a fraction of what he earned in IT. He's not alone. The Guardian profiled five skilled workers over 50 who've turned to data annotation after age discrimination locked them out of their fields. A former ER physician. A former academic. All with advanced degrees and years of experience, now labeling data for OpenAI, Google, and Meta through contracting firms like Mercor, TEKsystems, and Alignerr.
The pay range tells the story. Top experts with specialized knowledge can earn over $180 an hour, but most gigs pay $20 to $40, according to online job postings. No benefits. No stability. Contract work that could end anytime. Richard Johnson, vice-president of the AARP Public Policy Institute, told the Guardian that workers over 60 take roughly 50% longer to find new jobs than people in their 20s and 30s. Joanna Lahey, a professor at Texas A&M who studies age discrimination, calls these 'bridge jobs,' lower-paying roles that keep workers afloat near retirement. Historically that meant retail or driving for Uber. Now it's training the AI that might replace them.
The supply chain makes it worse. Intermediaries like Scale AI and Appen sit between tech giants and workers, sometimes taking 30 to 50% margins. Workers get classified as independent contractors, so companies avoid paying benefits or minimum wage. Critics say this structure insulates OpenAI and Google from direct responsibility for labor conditions. About half of US workers aged 50 to 54 get pushed out of long-term jobs before they expect to retire, per the Urban Institute. The pandemic made it worse. Roughly 5.7 million workers over 55 lost jobs in early 2020. 'There's just a lot of desperation out there,' Johnson said. Ciriello's situation is extreme but not unusual. Experienced professionals are teaching AI systems to do the very jobs they can't get hired for anymore.