New Concierge
The real reason why the high value shift is a lie (that is being told and repeated in India), is that there are no signs that we out there creating a nation of AI architects. Thus far, the focus has been on creating a nation of managed labor. For decades, the tech industry in India was built on execution. You took a task, basic Java, QA testing, something mind-numbingly routine, and you did it faster than a guy in a cubicle in Dayton. AI ended that. If you’re a 22 year old graduate looking for a starter desk job, you’re competing with an AI that costs less than the average person's monthly Chinese takeout budget and does not sleep. But the jobs haven't vanished into the ether. They just moved back to the physical world where machines can’t reach yet. Snabbit crossed a million jobs by focusing on cooks and house help. They realized that while an AI can write a legal brief, it can't scrub a bathroom floor.
We just need to look at where the capital is actually flowing to separate fact from fiction. It’s not just Snabbit. There is Krutrim, Ola’s pet unicorn, trying to build the full stack of cloud and models, and Sarvam AI chasing the multilingual dream. These companies are building the brains, but the bodies they manage are the next million jobs. None of them look like the up-leveled jobs that the political and technology leadership in India are touting. Then there is M, the new startup from the Dunzo co-founder. It’s a concierge platform designed to automate household tasks through software coordination. These founders aren't building for the software engineer in Bangalore. They’re building for the physical service provider who is safe from automation because a robot that can fold laundry costs more than a house in the Mission District.
The shift is in introducing AI-driven surveillance of the work that was already being done. Snabbit uses Kavach, an AI that tracks voice decibels and phone patterns to protect workers. This is high-tech intrusion rebranded as a safety shield. The high value role here belongs to the platform owners, not the people on the ground. The AI does the scheduling, the quality control, and the monitoring. The human provides the muscle. It’s a million jobs, sure.
The big risk for major GCCs in Bangalore is the senior only trap. These firms want experts to audit the AI. They don't want to hire juniors because AI is now the junior. If you don't hire 20 year olds to do the boring work now, you won't have 30 year old experts in a decade. The career ladder is being hollowed out. You're either a member of the elite 5 percent orchestrating the machines, or you're working for an app in the physical service sector. The middle is evaporating and it seems like the best bet is to focus on physical complexity. Trust is the only thing that hasn't been commoditized. If a job requires a physical presence and a high degree of in-person human trust, it’s safe for now. Since AI can't go to jail, compliance and auditing roles are growing because a human still has to sign their name at the bottom. The trick is to get into this space, leverage it well and not turn into a gig worker for the likes of Snabbit.