Key Highlights
- Mercor, a San Francisco-based startup, is challenging the notion that human judgment is off-limits to artificial intelligence (AI).
- The company’s mission is to bridge the gap between machine learning and human nuance by training AI with core skills.
- Co-founders Adarsh Hiremath and Surya Midha, aged 22, were recently named the youngest self-made billionaires after their company was valued at $10 billion last November.
- Mercor’s AI Productivity Index benchmarks models on real-world knowledge in fields such as medicine, management consulting, investment banking, and law.
AI’s Limitations and Human Nuance
The debate over the limitations of artificial intelligence (AI) has gained significant traction with leaders like Jamie Dimon, CEO of JP Morgan, emphasizing emotional quotient (EQ) and critical thinking as indispensable skills. These skills are seen as the last bastion against AI’s encroachment on human work domains. However, a young Silicon Valley startup is pushing back against this narrative.
Mercor: Bridging the Human-AI Gap
Mercor, founded by 22-year-old billionaires Brendan Foody, Adarsh Hiremath, and Surya Midha, aims to democratize AI training. The company’s unique approach involves hiring experts from diverse professional backgrounds to teach AI systems human skills such as judgment, nuance, and taste.
“Everyone’s been focused on what models can do,” Foody told Fortune in November 2025. “But the real opportunity is teaching them what only humans know—judgment, nuance, and taste.”
The Challenge of Human Skills in AI
In a volatile labor market where even professional skills may not guarantee job security, the demand for specialized knowledge and human-in-the-loop expertise has surged. According to the World Economic Forum’s 2025 Future of Jobs Report, employers predict that 39% of core skills such as problem-solving and communication will be disrupted by 2030.
The stakes are high, with some AI thought leaders warning that half of all entry-level white-collar jobs could vanish soon. Despite these warnings, Mercor is dedicated to proving the executives wrong—by hastening the replacement of people with AI in the workforce and closing what they see as the last mile of human employment.
Contract Work and AI Training
Mercor’s career page lists numerous job postings for contract work across fields such as investment banking, private equity analysis, linguistics, sports journalism, and legal expertise. These roles offer hourly rates ranging from $10 to $150 per hour, with fully remote working conditions.
The company boasts an average hourly rate of $86, paying out about $2 million daily to its experts. To apply, candidates submit an initial application followed by an AI interview tailored to their area of expertise, reviewed by Mercor staff. Once hired, contractors evaluate the AI’s performance on micro-tasks using detailed rubrics.
The Quest for Nuance and Judgment
Mercor has launched its AI Productivity Index (Apex), which benchmarks models in real-world knowledge across medicine, management consulting, investment banking, and law. Even the most advanced models, like GPT-5, fail to meet the production bar, achieving a top score of 64.2%.
Despite these limitations, Mercor believes that AI performing at 60% or better can reshape work dynamics as professionals collaborate with technology. “Perhaps a consultant can more easily complete a competitor analysis if given an initial draft from an AI,” the company stated.
The Future of Work
As AI continues to evolve, the focus may shift from simply doing the work to possessing the judgment required to critique it. The 2026 Fortune Workplace Innovation Summit in Atlanta will explore how AI, humanity, and strategy converge to redefine the future of work.
Mercor’s vision is clear: by teaching AI human skills, they aim to bridge the gap between machine learning and real-world application. The race against time continues as the world watches closely to see if these young entrepreneurs can indeed prove their point.