Teach What AI Can't Automate – Moravec's Paradox Is Your Curriculum

01 /The Concept

One headline about Emversity puts the startup's mission bluntly: it helps companies "scale the number of employees AI can't replace" – by training people in exactly those disciplines.

Riding that narrative, the Indian startup doubled its valuation from its previous round in spring 2025 (when it was valued at $60 million) to $120 million on its latest $30 million raise.

Since its founding in 2023, Emversity has graduated just 3,000 students – which would look thin for a $120 million valuation, except that the startup is training people for roles the market desperately needs and AI can't easily fill.

Currently, Emversity operates in two sectors: healthcare (nurses, radiology technicians, lab technicians, elder care) and hospitality (hotel front-desk staff).

With the new capital, the startup plans to expand into engineering, procurement, construction, and manufacturing.

Why healthcare? It's one of the fastest-growing sectors – and understaffed by an estimated 20% of positions, a gap projected to worsen as global demand rises. By 2030, the sector is expected to need 97 million workers worldwide.

Why hospitality? The market is expanding, driven partly by domestic tourism. In 2024 alone, 1.7 billion domestic trips were recorded in India. The sector is expected to generate 50 million new positions in the next 5–7 years, on top of existing roles.

Emversity's base model is B2B: it embeds its curriculum inside universities and charges the universities, which benefit from improved graduate employment rates.

Additional revenue – representing 20–30% of the total – comes from career counseling services for students and short certification prep courses.

The founders report 80% gross margins and a student acquisition cost below 10% of revenue, citing the university channel as their primary organic acquisition engine.

02 /Why It Matters

Emversity focuses on sectors where the work is skill-based rather than knowledge-based – where you need to be able to do something with your hands, not recite theory.

The startup also localizes its programs by geography, placing different curricula in different regions based on local demand: healthcare programs dominate in southern markets, while hospitality training is more relevant in the northeast.

The focus on physical skills is a direct application of Moravec's Paradox, first articulated in 1988. The paradox: high-level cognitive tasks – chess, calculus, language translation – are relatively easy to automate with modest compute. But low-level sensorimotor skills – perceiving the physical world, using your hands, adapting in real time – require enormous computational power and remain extremely hard for AI.

Put differently: what we consider difficult is easy to automate. What we consider simple is actually very hard to automate. Don't expect robot nurses or robot plumbers anytime soon. And by extension, people who perform skilled physical work have a much longer runway before they need to worry about AI competition.

The skills shortage in these trades compounds the opportunity. At least 25% of the current workforce in these professions is nearing retirement – and there aren't enough replacements coming in. For every five workers who leave, only two new ones enter. Younger people avoid these careers despite the fact that skilled tradespeople often earn more than most office workers.

That creates a dual challenge: not just training people, but attracting them to careers they currently overlook. Changing the perception of these professions – among potential candidates and in the broader culture – will require deliberate effort.

03 /Opportunities

For years, education startups focused on teaching technology – coding, then AI. That's still crowded and increasingly commoditized.

The more interesting market is hiding in plain sight: training for skilled-trade and hands-on professions. This is where the labor shortage is already acute, where the work is structurally hard for AI to replace, and where employers are consistently underserved.

The opportunity: build education programs for these fields. It may not sound glamorous – but the demand is undeniably real.

The critical insight from Emversity's model is that writing the curriculum is actually the easy part. The hard problems are distribution (how do you get the right candidates into your program affordably and reliably?) and placement (how do you build consistent relationships with employers who will hire your graduates?).

Solve those two problems for a specific trade in a specific region, and you've built something durable.

So: who would you train, where would you find them, and who would hire them?