When the IT dream falters

It was a moment of ironic clarity when Geoffrey Hinton, the ‘godfather of AI’, suggested, half-jokingly, that young people may be better off becoming plumbers than tech engineers. On the surface, it sounded absurd. But beneath the quip lies a cold, clear warning: in a world where AI is eating into digital jobs, physical, hands-on work might offer more security than an entry-level job writing Python scripts. This wasn’t just about plumbing; it was a wake-up call to a generation still riding the ghost of the 2000s IT boom.

Two decades ago, Indian engineering grads didn’t just enter the workforce; they were part of a national phenomenon. IT jobs were more than employment opportunities; they were social escalators, offering economic dignity and foreign visas. I still remember reporting from a grand IT service firm’s event in 2004 at the Taj Mumbai when it had just crossed a billion dollars in revenue. That evening glittered with promise. Cousins were flying off to the US with a little more than some C++ and Oracle skills. It felt like the digital revolution was India’s to command.

Advertisement

Fast forward to 2025. That escalator is stalling, if not shutting down entirely. Entry-level jobs, the same ones that launched lakhs of careers, are being rendered obsolete not by mismanagement or recession, but by machine learning and automation. AI has rewritten the script.

The hiring numbers tell the story. Wipro, which welcomed 38,000 freshers just two years ago, is now down to 10,000. TCS added only 625 people last quarter. Infosys has delayed onboarding for over a year, leaving thousands of young graduates in limbo. The conveyor belt that once pulled young Indians into middle-class stability is now rattling to a halt. It’s not because the youth are less smart. It’s because AI is faster, cheaper, and increasingly capable of doing the jobs humans used to do.

From GitHub Copilot writing boilerplate code to ChatGPT assisting with debugging, the very definition of a ‘tech job’ is changing. The tools that helped engineers are now taking their place.

This is not just a structural crisis. It’s an emotional one. Young people across India, particularly those from tier-2 and tier-3 colleges, find themselves abandoned in a job market they were told they were being prepared for. The India Skills Report 2025 states that only 47 per cent of engineering graduates are employable. The rest, armed with degrees but no direction, are stuck in PG rooms across Bengaluru, Pune, and Hyderabad, refreshing job portals, chasing internships, and chasing hope. For many, this isn’t about upskilling; it’s about identity. They were supposed to be engineers. Instead, they’re now upskilling for gig work they never imagined doing.

This isn’t just happening in India. In the US, the unemployment rate among computer science graduates is nearly double the national average. In the UK, some graduates are applying for a thousand jobs without landing a single interview. In global boardrooms, the consensus is brutal: entry-level roles are the easiest to automate. According to the World Economic Forum’s Future of Jobs Report 2025, 40 per cent of employers expect to reduce headcount due to AI. While 170 million jobs might be created this decade, 75 million will disappear, mostly those involving repeatable, low-risk tasks.

In the middle of this churn, Global Capability Centres (GCCs), the new hubs for firms like Goldman Sachs, Walmart, and Siemens, are hiring, but very selectively. They’re picking from NITs, IITs, and IIITs. They want thinkers, not testers. Internships are their filters. If you don’t intern, you don’t get in. The old system of being trained after being hired is gone. Today, you prove your value before you enter the door. The gap between the elite and the average is widening.

The academic institutions, meanwhile, seem stuck in time. A staggering 93 per cent of Indian engineering colleges don’t offer full AI or data science coursework. Only 7 per cent of faculty have worked with generative AI. We’re teaching 2015 in 2025. Syntax-heavy syllabi and handwritten exams still dominate even as the industry demands systems thinking, ethics, domain knowledge, and experience with agentic AI tools. The real tragedy is that this disconnect isn’t just technical, it’s existential. Families are spending lakhs on a B.Tech degree, imagining it to be a ticket to upward mobility. They’re not just paying for a degree; they’re paying for a dream. And increasingly, the dream is hollow.

Still, there’s a narrow but growing path forward. AI isn’t just killing jobs; it’s also making new ones in areas few institutions are preparing students for. Roles like prompt engineers, AI ethics testers, model debuggers, data labellers, and chatbot QA testers are becoming mainstream. These don’t necessarily require four-year degrees. What they demand is curiosity, adaptability, and tool fluency.

We must radically reimagine our tech education. AI and data science must be core parts of engineering education, not side electives. Exams must give way to hackathons. Curriculum must shift from rote learning to building and testing real-world solutions. Degree-plus-internship models should become standard. Let students earn, fail, and rebuild, not just cram for tests.

For students and young professionals, the message is simple, if not easy. The old career ladder is broken. But there’s a new highway forming, one filled with short-term certifications, self-led learning, and stackable micro-skills. You don’t need to be in Silicon Valley or have a Stanford degree. You just need to stay updated, stay nimble, and stay hungry. AI doesn’t have to be your competitor. It can be your co-pilot.

And finally, yes, take Hinton’s ‘plumbing’ line seriously. He wasn’t mocking tech. He was reminding us of a fundamental truth: jobs that exist in the physical, practical world, jobs requiring real-world judgment, touch, and repair, are incredibly hard to automate. Maybe you won’t be a plumber. But maybe you’ll be an AI hardware technician. Or a drone repair expert. Or a robotic arm calibrator. These are the new hands-on jobs, digital yet physically rooted in reality and often beyond the reach of pure automation.

The IT dream isn’t dead. But it’s no longer a dream for the masses. It’s become selective, elite, and redefined by AI. The real opportunity now lies not in chasing yesterday’s jobs but in preparing for tomorrow’s roles, even if they don’t have the glamour of a ‘software engineer’ tag. Because in the new world of work, glamour doesn’t pay the rent. Skills do. Relevance does. Reality does. Hold on. The future is still yours, but only if you’re ready to work for it in new, sometimes unexpected ways.

The writer is a techie, political analyst, and author.