Mercor, a three-year-old startup, has become a $10 billion intermediary in the AI data gold rush. The company connects AI labs like OpenAI and Anthropic with former employees from Goldman Sachs, McKinsey, and big law firms to share industry expertise and pay them up to $200 an hour to train AI models that can ultimately automate their former employers out of business.
Today we’re bringing you a conversation with CEO Brendan Foody from this year’s Disrupt. There, he explained why AI labs need highly skilled contractors rather than a crowdsourced workforce, how Scale AI’s problems accelerated Mercor’s rise, and why he thinks the entire economy will focus on training AI agents.
Listen to the full episode to hear:
How Foody grew from high school AWS credit consulting to a $10 billion valuation Why the top 10-20% of contractors are driving the majority of model improvements and how Mercor finds them The gray area between employee knowledge and trade secrets (and whether Goldman Sachs should worry) Why Foody believes all knowledge work will eventually become training data for AI agents
Subscribe to Equity on YouTube, Apple Podcasts, Overcast, Spotify, and all casts. You can also follow Equity on X and Threads at @EquityPod.
