San Francisco, October 5 — Mercor, the San Francisco-based AI training company, has grown rapidly since its founding four years ago, expanding from a small team to a multibillion-dollar enterprise that now engages over 100,000 freelancers worldwide to help train artificial intelligence models.
The company’s rise was highlighted in a recent 60 Minutes segment, where CEO Brendan Foody described how Mercor began with just a handful of people after the release of ChatGPT, when he and two high school friends recognized that AI systems needed more specialized training data than what was freely available online.
Foody, who is 23 years old and described as a billionaire, explained that Mercor now employs people from a wide range of professions — including winemakers, music producers, physicians, poets, architects, and finance professionals — to provide the nuanced, real-world expertise that AI lacks. These freelancers typically work on AI training as a side gig while maintaining their full-time careers. One example cited was Shawna Miller, a winemaker who contributes her sensory and industry knowledge to help train models related to viticulture.
Another was Robbie Hiser, a music producer, who argued that AI cannot replicate certain human experiences — such as going through a romantic breakup — and therefore cannot fully capture the emotional depth involved in songwriting. Dr. Melania Poonacha, a physician, was also mentioned as someone who applies her medical expertise to AI training efforts outside her clinical work.
Foody emphasized that the goal of Mercor’s approach is not to replace human workers but to offload routine tasks to AI, allowing people to focus on higher-value work that requires judgment, creativity, and lived experience. He suggested that as AI handles more repetitive functions, demand for human expertise in areas like subjective evaluation and complex decision-making could actually increase, rather than decline.
This view contrasts with recent layoffs in the U.S., where companies have cited AI as a reason for workforce reductions, though the 60 Minutes transcript did not specify how many jobs were affected or which companies were involved. The discussion on the program centered on whether AI will reshape work in the way past technological shifts have — transforming roles without eliminating the need for human input — or whether it might lead to widespread job displacement.
Foody maintained that the evolution of work is a natural process and expressed confidence that humans will continue to perform tasks that rely on personal experience and subjective judgment, which machines cannot replicate. Looking ahead, the growing reliance on diverse professionals to train AI systems raises questions about how such collaborations might shape the future of work across industries.
As more individuals from fields like medicine, the arts, and skilled trades contribute their knowledge to AI development, the boundary between human expertise and machine learning may continue to blur — not to replace people, but to redefine where their unique strengths are most needed. For now, Mercor’s model illustrates one path forward: a future where AI learns from humans, not instead of them.
