Artificial intelligence (AI) is changing many jobs, even medicine. A great example is Dr. Alice Chiao’s case, a physician who used to teach Emergency Medicine at the Stanford University School of Medicine and now spends part of her time training artificial intelligence systems to respond the way a real doctor would.

So, let’s find out more about how this process works, why this has become a millionaire business, and what it really means for the future of this job. Let’s get started.

What does Dr. Alice Chiao really do?

Dr. Chiao is currently training artificial intelligence models how to think, diagnose, and respond as a doctor. The process she uses is called ‘’reinforcement learning,’’ which means that artificial intelligence gives an answer to a question, and a human expert evaluates it. If the answer is right, safe, and clear, it will receive a good rating. However, if the answer is confusing, wrong, or potentially dangerous, it will be corrected. With the passage of time, the system learns how to improve because it’s designed to try to obtain better evaluations.

Chiao uses real situations she has experienced as a doctor in both primary care and emergency medicine. So, she can ask questions from a patient’s point of view, for example, if a child has fever, should he go to the doctor? She also checks how the system responds to medical terminology, like what appears on hospital intake forms.

Artificial intelligence business

Dr. Chiao works with a company called Mercor, which connects experts from different fields with companies developing artificial intelligence. Mercor has hired professionals from fields like medicine, law, finances, sports, comedy, and even the wine industry. These experts can earn hundreds of dollars per hour by evaluating and correcting responses coming from artificial intelligence systems.

According to Dimitri Zabelin, a senior AI analyst at PitchBook, this type of AI training industry is estimated to be worth at least $17 billion.

Major AI companies such as OpenAI, Google, and Anthropic rely on this process. Mercor’s CEO, Brendan Foody, described it as using “large armies of people” to train models properly. AI systems may be trained on huge amounts of data, but without human feedback, they do not improve effectively.

Will artificial intelligence replace human professionals?

One of the main concerns is whether artificial intelligence will replace human jobs. Some people believe that companies like Mercor are making this situation easier by helping AI systems become good enough to eventually replace stable, full-time careers. So, instead of permanent jobs, workers may end up doing temporary contract work training machines that will later automate their roles.

However, Dr. Chiao doesn’t agree with this. She said she is not doing this process to make people lose their jobs, but to help doctors because artificial intelligence could eventually assist doctors by reading scans, filling out medical records, or taking notes during patient visits. These tasks often consume large amounts of time and if AI handles administrative duties, doctors could spend more time speaking directly with patients.

At the same time, she emphasizes that AI is not a doctor. It does not have intuition. It cannot look a patient in the eyes and sense something beyond test results and written symptoms. She recommends that patients use AI tools as a starting point for information, not as a replacement for a trained physician with years of experience.

The growth of Mercor and other companies

Mercor was founded three years ago by Brendan Foody, Adarsh Hiremath, and Surya Midha. At first, it was a platform to help people find jobs, but it changed its focus to artificial intelligence.

Foody stated that the company pays more than $1 million per day to thousands of experts. So, in less than two years, the company reportedly grew from projecting $1 million in revenue to over $500 million. According to Zabelin, Mercor is valued at more than $10 billion.

Mercor is not alone in this field. Last year, Meta invested $14 billion in Scale AI, which operates in a similar sector. Investors believe human feedback and expert model testing will remain a permanent and essential part of artificial intelligence development.

So…

So as artificial intelligence continues to evolve, the key question is not only what AI can do, but how humans choose to shape it. In this case, experienced professionals are working to ensure that technology supports their work rather than replaces it. And that decision — how we guide and use artificial intelligence — will influence the future of many industries, including healthcare.