Utkarsh Amitabh wasn’t looking for a new job in January 2025, but when a startup reached out to him to join its network of human experts for AI training, intellectual curiosity drew him in. Now, he earns $200 an hour in a side gig that combines his passion and income.
Article contents:
- An unexpected opportunity
- The expert network and the human role
- Nature of the work: Accuracy and detail
- Financial reward and intellectual challenge
- The future of jobs and artificial intelligence
- Frequently asked questions
An unexpected opportunity
Amitabh, based in the United Kingdom, agreed to take on a part-time freelance role training AI models for the startup “Micro 1”. Amitabh has a busy schedule as an author, university lecturer, and entrepreneur, in addition to pursuing his Ph.D. However, he found that the opportunity to train AI models fit perfectly with his diverse background in engineering, philosophy, and business, making him the ideal candidate for this detail-oriented role.
The expert network and the human role
The company has built a network of over two million experts working to train AI models for major clients. The company’s marketing director says that current models have absorbed most general knowledge, and that real progress now comes from specialized experts who can challenge and refine the model. The evolution of artificial intelligence relies heavily on this high-quality human input to ensure accuracy and reliability.
Nature of the work: Accuracy and detail
The work involves feeding the model complex information and scenarios, then improving it through testing. Amitabh often works on confidential business problems, breaking down complex issues into small pieces in clear language that the machine understands. If the model makes a mistake, he steps in to identify and address the error, a trial-and-error process that can take hours and requires immense attention to detail to catch subtle mistakes.
Financial reward and intellectual challenge
Amitabh now makes $200 an hour, and has earned nearly $300,000 since January. He describes the job as intellectually demanding because the models are constantly learning, forcing him to elevate his own knowledge. He views the ultimate goal as highly motivating, as he witnesses how human-machine collaboration can elevate problem-solving.
The future of jobs and artificial intelligence
Amid fears of human replacement, Amitabh describes himself as a pragmatic optimist. He acknowledges challenges and some job losses, but is confident that knowledge is not limited, and that humans and machines will always share a symbiotic relationship. He believes this type of work proves that human expertise will remain a rare and valuable currency in the era of automation.
Frequently asked questions
Question: What are the qualifications required for this work?
Answer: Companies typically require deep expertise and specialization in fields such as law, medicine, programming, or business, in addition to critical thinking skills.
Question: Can anyone do it?
Answer: Not easily. Companies look for “experts” with very specialized knowledge to train models on complex tasks that public data cannot solve.
Question: Is this work sustainable?
Answer: Yes, as long as AI models need to evolve and learn, the need for human expertise to evaluate and guide them will remain.