Imagine turning your intellectual passions into a lucrative side hustle that pays $200 per hour—while juggling a newborn, a Ph.D., and running your own company. That's the reality for 34-year-old entrepreneur Utkarsh Amitabh, who dove into training AI models for a data labeling startup called micro1. But here's where it gets intriguing: Is this the future of work, or a stepping stone to our own obsolescence in a world dominated by machines? Stick around to uncover his story and the bigger questions it raises about AI's role in our lives.
Back in January 2025, Utkarsh wasn't hunting for a new gig. He was already swamped with responsibilities: penning books, lecturing at universities, leading Network Capital—a global platform for mentorship and career growth—as its founder and CEO, and pursuing his doctorate at the University of Oxford's Saïd Business School. Oh, and caring for his brand-new baby at home, as he shared with CNBC Make It. Yet, when micro1 reached out, inviting him to join their team of human specialists who refine artificial intelligence systems for major companies, his curiosity won out.
'Intellectual curiosity drew me in,' Utkarsh admits. The idea of shaping enterprise-level AI models resonated deeply with his expertise in business strategy, financial modeling, and technology. Micro1, you see, seeks out seasoned professionals with specialized knowledge—from physicians and attorneys to engineers—creating a pool of 'deep generalists' who can handle diverse challenges. Utkarsh fits that mold perfectly, thanks to his mechanical engineering bachelor's, master's in moral philosophy, and over six years at Microsoft in business development, where he focused on cloud computing and AI collaborations. His published works even include a book on 'the side-hustle revolution' and a thesis exploring how AI might redefine human achievement.
This role felt like a seamless extension of his interests, he explains. Plus, the part-time, freelance setup offered flexibility—he typically puts in about 3.5 hours nightly, after his 1-year-old daughter is asleep. 'This didn't seem like an add-on, but something that I could use to further my interests in a limited number of hours a week,' Utkarsh says. And the payoff? He earns $200 an hour training these AI models, confirmed by a pay stub reviewed by CNBC Make It. A micro1 spokesperson also verified that, including bonuses for completed projects, he's pocketed nearly $300,000 since starting in January.
That said, financial gain wasn't his primary driver. With a steady income from his other ventures, Utkarsh emphasizes that 'money was less of a motivator' than the alignment with his passions. Still, he values fair compensation, calling the pay 'respectable' for such specialized work requiring deep expertise.
Now, let's break down what this job entails, especially for those new to AI. Founded in 2022, micro1 has assembled a vast network of over 2 million experts to train AI systems for clients like big AI labs (think Microsoft) and Fortune 100 companies building custom large language models for their teams. As per TechCrunch, micro1 recently hit a $500 million valuation and competes with players like Mercor and ScaleAI.
These experts, including Utkarsh, form the 'backbone of our data quality,' according to micro1's chief marketing officer, Daniel Warner. He explains that modern AI has already consumed most public knowledge, so progress hinges on domain specialists who challenge, refine, and outsmart the models. The 'human data' they produce delivers top-tier results for elite AI labs and major corporations.
So, how does AI training work? Simply put, it involves feeding algorithms with enormous datasets of information and scenarios. The model then gets fine-tuned through testing with prompts, like asking an AI to manage expenses, forecast growth, or draft a budget for a business division. For beginners, think of it as teaching a computer to think step-by-step, much like guiding a student through complex math problems.
Utkarsh's projects often remain confidential, focusing on dissecting intricate business dilemmas that everyday users, business owners, or executives might face, breaking them into bite-sized pieces. This mirrors prompt engineering—a hot skill paying over $100,000 annually, as noted in another CNBC piece—where you craft precise, machine-friendly language to elicit accurate responses from AI.
If the AI's output misses the mark or veers off-track, Utkarsh spots the glitches, like overlooked details or lost nuances, and adjusts the dataset. It's an iterative process, sometimes spanning hours per problem set. 'You need to have immense attention to detail, and you have to often look out for mistakes that the human might make or a machine might make, and you discover more about the kinds of mistakes that exist by the process of immersing yourself in it,' he notes.
The work is intellectually rigorous, too, because AI systems evolve rapidly, pushing experts like Utkarsh to update their own knowledge and creativity. 'The ultimate goal is actually really energizing,' he adds. 'You're seeing whether the machine and human, the way this engagement is happening, [can] level up the output for problems that you asked and other kinds of problems that might be related to it.'
And this is the part most people miss: Amid AI's workplace surge, concerns about job displacement loom large. Will advanced tech render human workers unnecessary, or drastically alter their roles? Utkarsh weighs in on 'the trillion-dollar question,' describing himself as a blend of techno-optimist and techno-realist.
He acknowledges the 'growing-up pains' as firms adopt AI, potentially wiping out jobs—an impact HR experts say is already underway, with surveys indicating 89% of roles could be affected next year. Yet, he's hopeful that AI will generate new opportunities to balance the losses. For example, a World Economic Forum report from January 2025 predicts AI's disruption could yield about 80 million net jobs by 2030, transforming the global market for the better.
Philosophically, Utkarsh believes knowledge isn't limited—humans and machines can collaborate symbiotically, propelling each other forward. 'It's also possible that this AI fear collectively empowers us to learn better, upskill ourselves and frame questions differently about ourselves,' he says. 'So I'm not concerned about the [idea of] AI Doom entirely, because I think it does far more good than bad.'
But here's where it gets controversial: Could training AI like Utkarsh is doing actually accelerate the very job losses he's optimistic about offsetting? Some might argue it's unethical to fine-tune machines that could replace experts, while others see it as inevitable progress. What do you think—does empowering AI with human expertise lead to a brighter, more collaborative future, or does it set the stage for widespread unemployment in specialized fields? Share your views in the comments below. Do you agree with Utkarsh's techno-realist stance, or lean more toward pessimism? We'd love to hear your take!
Ready to boost your own career prospects in this AI-driven world? Sign up for Smarter by CNBC Make It's online course, 'How to Build a Standout Personal Brand: Online, In Person, and At Work,' available here. Discover strategies to highlight your unique skills, cultivate a strong reputation, and establish a digital footprint that even AI can't duplicate.
