A profile of AI contractor marketplace Mercor, which is valued at $10B and whose three 22-year-old founders are Thiel Fellows, each holding a roughly 22% stake
“Mercor, a recruiting startup that helps Silicon Valley's biggest AI labs … Bluesky: Dare Obasanjo / @carnage4life : The founders of Mercor are the world's youngest self made billionaires at 22 as their startup is now worth $10B and they each own 22%. — It started off as a recruiting website but is now a place for AI companies to hire experts like doctors, lawyers and PhDs to train & evaluate AI model outputs.
Context & Ripple Effects
Mercor’s valuation marks a sharp step-up from the $2B valuation attached to its February fundraise. Related coverage had already positioned it as a Scale AI rival that recruits domain specialists to train and assess models.
The reported value also follows news that Mercor was finalizing a $350M round at the same $10B valuation, making the profile a window into who owns the upside from its contractor-marketplace model.
First-order effects
- At a $10B valuation, each founder’s roughly 22% holding carries an implied paper value of about $2.2B, concentrating most of Mercor’s equity with its founding team.
- Mercor gains additional visibility with AI labs and expert contractors as a marketplace for specialized model-training and evaluation work.
Second-order effects
- Competing AI-data and contractor platforms face greater pressure to secure scarce doctors, lawyers, and other domain experts, since expert supply is central to differentiated evaluation services.
- AI labs using external expert workforces gain a more prominent intermediary, potentially shifting recruiting and contractor-management activity away from direct lab operations.
Third-order effects
- If high valuations continue to accrue to marketplaces that organize expert feedback, human evaluation may become a distinct layer of AI infrastructure rather than a temporary extension of recruiting.
- The model could concentrate bargaining power in a small number of platforms that aggregate both AI-lab demand and specialized-worker supply; that outcome depends on whether labs keep outsourcing this work.
The trend: AI development is elevating platforms that turn specialized human judgment into an on-demand input for training and evaluating frontier models.