Mercor , and scientists are training artificial intelligence to do the jobs they once did. This paradoxical phenomenon reflects the rapid evolution of AI and its impact on the labor market, where traditionally high-paying office jobs are being automated by systems like ChatGPT.
See also: Microsoft: Hackers target developers with malicious Next.js repositories
The story of Katya, a freelance journalist who turned to content marketing only to discover that AI had automated much of the work, is typical. When she saw an ad from Crossing Hurdles promising copywriting jobs for $45 an hour, she initially dismissed it as a scam. But her desperation eventually led her to Mercor, where she realized she would be training the model who would replace her.
The hiring process at Mercor involves an interview with an AI chatbot named Melvin, which analyzes candidates’ resumes and asks targeted questions. After acceptance, employees sign contracts, undergo background checks, and install monitoring software on their computers. The work includes creating sample prompts, constructing ideal responses, and creating detailed evaluation criteria.
Mercor and the AI education ecosystem
Machine learning systems learn by spotting patterns in vast amounts of data, but first that data needs to be sorted, labeled, and produced by humans. ChatGPT gained its amazing fluency from thousands of people hired by companies like Scale AI and Surge AI to write examples.
See also: YouTube finally offers a proper alternative to ad-blockers
Outlier AI , one of Mercor ’s main competitors , has connected over 1,000,000 experts to AI training projects , offering them additional income while building resumes in prompt engineering . This trend reflects a broader shift in the job market, where 60% of jobs in 2018 didn’t exist 50 years ago, and AI is accelerating this cycle of new job creation.
However, working at Mercor is characterized by instability. Katya experienced the abrupt termination of her first project just two days after it started, with no warning or guarantee of continuation. This uncertainty is characteristic of the gig economy model, where workers are deprived of traditional employment protections.
Research from the Harvard Kennedy School shows that job training programs increase the earnings of displaced workers, but less so in sectors with high exposure to AI. Workers with an average income of about $40,000 a year see smaller benefits when their jobs are vulnerable to automation.
The phenomenon creates a feedback loop: people train AI to replace human jobs, accelerating the need for retraining. Federal Reserve shows that companies adopting AI prioritize retraining their staff.
See also: The official YouTube app on Apple Vision Pro
The future outlook includes the expansion of this gig economy as AI increasingly requires specialized human data for advanced models. Hybrid tools are expected to develop – AI copilots working with human coaches through workforce organizations. Challenges include training efficiency gaps in AI -heavy sectors , while success depends on ethical scaling of platforms like Outlier .
