Mercor Reaches $10 Billion Valuation by Hiring Specialists to Train AI on Office Work

Mercor from San Francisco has built a business on a paradox: the company hires qualified specialists to teach AI to perform their work. With this approach, a startup founded by college dropouts has grown to a $10 billion valuation and become a notable player in the race to automate office work.

How the approach works

To automate white-collar tasks, it is not enough to give a model access to the internet, documents, and corporate correspondence. You need live examples of how specialists make decisions, prioritize, fix errors, and bring a task to completion. In this scheme, Mercor acts as a bridge between human and machine: the company pays highly qualified workers for the transfer of practical skills that previously remained within the profession and were rarely formalized. Essentially, the startup converts human expertise into a training loop for AI.

If a model is to perform office work not at the demo level but in real-world processes, it needs not just correct answers but correct action logic. Therefore, the value here is not in abstract "neural network training" but in mass collection of quality professional patterns: how to analyze a task, which exceptions to notice, where to double-check conclusions, and how to format results so they can be used within a company. This layer of knowledge can be scaled, updated, and fine-tuned on new cases.

Why Mercor took off

For investors, this model looks understandable and scalable. Mercor is not just selling another AI assistant but operating at a more expensive and ambitious level—attempting to turn expert labor into a repeatable system for automating white-collar roles. The founders' story also amplifies interest: the startup was created by former students who dropped out of college, and in a short time they brought the company to a $10 billion valuation. On such a market, this reads as a signal: demand for intellectual labor automation is already large enough to justify huge stakes. In fact, Mercor is selling not a promise of a distant future but infrastructure to accelerate it right now.

• Access to professionals becomes raw material for companies training models
• Real work scenarios can be turned into data, and data into product
• Focus on office tasks opens a larger market than mass chatbots
• The more precisely AI mimics specialist expertise, the more noticeable the economic effect for business

Who this concerns

The most uncomfortable part of the story is that people are essentially helping create tools that over time may reduce demand for their own skills. For the employer this looks rational: first pay an expert for knowledge transfer, then use the model where a human was previously required. For the specialists themselves, the picture is more complex.

For now, such projects provide income and demand for competencies, but simultaneously push the market toward making part of intellectual work standardized, measurable, and thus easier to automate. Yet not all office work will disappear at the same pace. Where responsibility, context, client trust, negotiation, or non-standard solutions are needed, humans maintain a strong position.

But routine pieces of white-collar work—preparation of standard materials, preliminary analysis, information structuring, repetitive operations within processes—are increasingly viewed as suitable territory for AI. Mercor bets precisely on this transitional layer between expert humans and machines learning to reproduce their useful actions.

What this means

Mercor shows where the center of the AI race is shifting: from flashy demos to systematic capture of professional knowledge. If this approach scales, the main value will be not the models themselves but access to people who know how to turn their work into a set of solutions, rules, and actions intelligible to AI.