AI Training Boom: Mercor's $2B Revenue vs. Acemoglu's Unemployment Warning

Key points

• Mercor, an AI data company, has grown from 20 employees in January 2025 to 400, with annualized revenue surging from $1 million to $2 billion in 24 months, according to Axios.
• Mercor utilizes over 100,000 freelancers to train AI models in diverse fields like finance, poetry, and medicine, with gig rates ranging from $20 to $200 per hour, per CBS News.
• Nobel Prize-winning economist Daron Acemoglu warns that without intervention, unemployment could triple in the next decade, noting that the current pace of AI adoption is unprecedented compared to the Industrial Revolution, per CBS News.
• Stanford research indicates hiring is down for young people in AI-exposed jobs, such as software development, while U.S. Census data shows declining wages for recent college graduates in similar majors, per CBS News.
• Clara Shih, former AI executive at Salesforce and Meta, founded the New Work Foundation to help young workers navigate the shifting economy, emphasizing that AI impacts self-worth and community stability, not just job titles, per CBS News.

Background

This debate follows a September McKinsey report predicting 11 million U.S. workers will switch careers by 2035 due to AI. A recent Pew Research survey found that nearly three-fourths of Americans fear AI will take their jobs, a sentiment mirrored globally where 34 of 37 surveyed countries expect more job losses than gains. The U.S. government has also moved to establish an 'AI Force' and appoint an AI tsar to steer policy, reflecting a strategic priority in the U.S.-China tech rivalry.

How outlets are covering it

Outlets present divergent views on AI's impact. CBS News and Axios highlight Mercor's rapid growth and the optimistic view that AI training is a 'career of the future' that frees humans from dull tasks. In contrast, CBS News cites Acemoglu’s pessimistic forecast of potential mass unemployment, arguing that the number of AI trainers is small relative to those displaced. The Minnesota Spokesman-Recorder offers a distinct perspective through its new column 'Bridging the Code,' focusing on how AI reshapes workforce equity for Black professionals. It argues that AI can reproduce historical inequities through algorithmic bias and urges Black professionals to gain influence over AI decision-making rather than merely adapting to tools designed by others.

Why it matters

The rapid commercialization of AI training, exemplified by Mercor's revenue spike, contrasts sharply with rising anxiety among workers and economists. As AI automates routine tasks, the labor market faces a potential 'transition shock' where young workers and recent graduates are disproportionately affected. The debate centers on whether policy can shift AI from a tool of displacement to one of augmentation, preventing a scenario where automation outpaces human adaptability.

What to watch

Policy discussions are intensifying regarding tax codes that incentivize hiring humans over AI and regulations to ensure AI systems do not replicate bias. Companies are redesigning roles to focus on high-judgment tasks, while nonprofits like the New Work Foundation are emerging to support displaced workers. The window for proactive government and corporate intervention is described as closing, with experts urging immediate action to manage the pace of automation.