AI Tokens Overtake Salaries: Mercor's Expensive Bet
Mercor's CEO reveals the startup now spends more on AI tokens than employee salaries. Is this the future for corporate budgeting?
Mercor's CEO Brendan Foody isn't shy about the fact that his $10 billion startup is pouring more cash into AI tokens than paying human salaries. In a chat on the '20VC' podcast, Foody shared that Mercor's spending on AI tokens now exceeds what they pay their employees. That's not a throwaway statement. It's a sign of things to come.
The reality is, Mercor, a company helping OpenAI and Anthropic train AI models, is betting big on digital over flesh and blood. With around 300 employees as of October 2025, according to PitchBook, it's clear where their priorities lie. They didn't respond to a request for comment, but actions speak louder than press releases, right?
AI's Pricey Future
Mercor's approach isn't just an outlier. Foody believes it's a harbinger for corporate America. His prediction? In five years, the average company will spend more on AI compute than on payroll. Wild? Maybe. But not impossible. The company uses AI for everything from project management to fraud detection, with over 5 million AI-assisted interviews under their belt. Are they paving the way or digging a hole?
Where's the ROI?
Not everyone is convinced about this AI spending spree. Uber's COO, Andrew Macdonald, recently expressed skepticism, noting a lack of clear productivity gains despite rising AI costs. Foody argues that lower costs and better models are creating a Jevons paradox, where instead of saving money, companies just use more AI. It's a classic case of tech optimism clashing with real-world results.
So, does this mean AI is set to become a bigger line item than employee salaries? Foody thinks so. He envisions a world where AI is as essential as electricity, outstripping labor costs. But here's the kicker: humans won't be out of the picture entirely. They'll still handle what models can't.
The question isn't if AI will dominate budgets, but when and how effectively. Are companies ready for it? Will this spending translate to actual value? Show me the product, not just the bill.
Get AI news in your inbox
Daily digest of what matters in AI.