Here's something to think about: while everyone's worried about AI replacing human jobs, one company is spending more than $1.5 million every single day paying people to make AI better at those jobs.
Mercor, a San Francisco startup now valued at $10 billion, has quietly built what CEO and cofounder Brendan Foody describes as "a new category of work." In a recent company blog post, Foody explained that the future isn't about AI eliminating human expertise—it's about humans teaching machines the subtle skills that only we possess, at least for now.
The Business of Teaching Machines to Think
Foody appeared on "The TBPN Show" podcast last month, telling hosts John Coogan and Jordi Hays that Mercor has been "growing like crazy" since announcing its valuation. The company now works with over 30,000 contractors worldwide who specialize in training AI models across diverse fields including software engineering, banking, and law.
The concept is straightforward but powerful. Instead of doing repetitive work forever, these contractors teach AI agents how to handle tasks once, enabling those systems to replicate the work millions of times. As Foody wrote, millions of people will "spend the next decade teaching machines the judgment, nuance, and taste that only humans possess."
When Training AI Pays Better Than Many Jobs
The economics are compelling. Contractors can earn up to $100 an hour teaching chatbots how to understand tone, context, and cultural nuance. Some trainers work with systems like xAI's Grok, helping them grasp internet culture and conversational patterns. Others focus on domain-specific expertise, fine-tuning model performance in specialized areas.
This isn't just a lucrative gig for contractors—it's created an entire ecosystem of billion-dollar companies. Scale AI and Surge AI, both startups connecting human trainers with AI labs, have achieved multibillion-dollar valuations that turned their founders into some of tech's newest billionaires.
The wealth creation has been staggering. Surge AI CEO Edwin Chen is worth approximately $18 billion according to Forbes estimates, while Scale AI co-founders Alexandr Wang and Lucy Guo have net worths of roughly $3.2 billion and $1.4 billion, respectively.
Going Public
When asked about Mercor's future, Foody told the podcast hosts that an IPO is "potentially on the horizon," though he stopped short of providing a timeline. Given the company's trajectory and investor appetite for AI-adjacent businesses, going public seems like a natural next step.
What Mercor's growth really demonstrates is that despite all the panic about artificial intelligence replacing workers, humans remain absolutely critical to making these systems function properly. The machines need us to teach them the soft skills, the judgment calls, the cultural awareness that can't simply be programmed.
As Foody framed it in his blog post, we're entering an era where people teach machines the same creativity and discernment that once defined human work itself. It's an interesting twist: AI doesn't eliminate the need for human expertise—it just changes how we apply it.