AI Isn’t Taking All the Jobs, But It Is Changing Who Gets One

Shantanu Chaudhary

Everyone is certain AI is changing the job market, that it is coming for their jobs. The data says something stranger—and more useful—than either the panic or the comfort.

Somewhere in a town that is not a metro—Kanpur, say, or Kochi, or Kota—a young man in his final year of engineering refreshes a job portal for the ninth time before lunch. He has done everything the script asked of him. The coaching classes, the ranks, the projects uploaded to a profile nobody seems to open. His parents borrowed against a plot of land to get him here.

On the family WhatsApp group, an uncle has just forwarded a video with a scarlet thumbnail: AI WILL DESTROY 50% OF JOBS. Below that, a cousin has typed sach mein? Is it true? The young man doesn’t reply, because he has been refreshing the same portal for weeks, and the listings that used to appear each morning have thinned to almost nothing, and lately he has begun to wonder the same thing himself.

Here is the strange part—the part neither the doom video nor the soothing LinkedIn post will tell him. Fear and fact have drifted apart, and the distance between them is dangerous. The panic, right now, is far larger than the numbers justify. And the comfort that every new technology creates more jobs than it destroys, is quietly wrong in exactly the spot where this young man happens to be standing.

The Reassuring Averages vs. The Real Fear

Let’s start with what is genuinely reassuring, because it is real. The economy-wide catastrophe many predicted has not materialised. The aggregate figures are calm; the great tsunami of AI layoffs that was promised did not arrive as a tsunami.

History is on comfort’s side, too. For two centuries, every machine that was supposed to end human work instead shoved it somewhere new, and the men who smashed the looms and, later, the ones who feared the computer were, in the end, wrong. On paper, even the largest forecasts still show more roles created than destroyed by the close of this decade. If all you read were the headline numbers, you would tell the boy at the portal to calm down and keep applying.

But the fear is not baseless, and it is worth asking why it runs so far ahead of the data. It is partly because we are built to feel a vivid story more than a dull statistic. The friend who didn’t get placed, the red-thumbnail video, the cousin’s one-word dread all land harder than any reassuring national average.

And partly because, underneath the noise, people can sense something the averages haven’t caught up to yet. This particular machine is not replacing the muscle, the way the last two centuries of machines did. It is coming for the desk. For the clean, air-conditioned, English-speaking work that an entire generation was told was the safe harbour. When the harbour itself starts to look exposed, no amount of calm arithmetic quite settles the stomach.

AI Changing the Job Market

And the stomach is onto something, because AI is changing the job market, the damage is real—it is simply not landing where layoffs land. It is landing at the entrance. The junior analyst, the first-year associate, the fresher writing basic code or closing a simple support ticket. That is precisely the rung a capable machine now covers.

And here is where the fear and the facts genuinely tangle, because the softening at the entrance does show up—in fewer freshers hired, in thinner campus placements, in the quiet swell of graduates who did everything right and are still waiting.

The catch is that we already keep a shelf of explanations for those numbers. Too many colleges, a slow hiring cycle, degrees that don’t match the market. So a genuinely new cause slips in wearing the old causes’ clothes.

The overall jobless rate barely stirs; the graduate’s has been troubling for years, troubling enough that one more reason for it hardly announces itself. It takes a narrower, colder look to pull the new thread out of the old tangle. And that look has now been taken. Over the past two years, employment for the youngest workers in the most exposed roles has begun to slip in step with that exposure, even as everyone insisted nothing had changed.

India’s Unique Vulnerability

Here the story turns specifically, uncomfortably Indian. The miracle that built our urban middle class. The outsourcing boom, the glass towers of Bengaluru and Hyderabad and Gurugram, the headset-and-cubicle jobs that lifted millions of ordinary graduates into salaries their parents could not have dreamed of.

It rested on one thing above all: English-speaking humans doing repetitive, rules-based, scripted work at enormous scale. That is not a flattering description of India’s great advantage. It is an almost exact description of what artificial intelligence does best.

Our own government’s planners have said as much, warning that in the worst case scenario, the country’s tech-services headcount could shrink by something like a fifth to a quarter by the early 2030s. The most AI-exposed category of work on the planet is the one we made our national fortune performing.

The Cruel Trick of “Job Creation”

There is another cruel trick hiding inside comforting averages. The numbers might proudly declare that for every job lost, a new one is created, keeping the national balance sheet steady. What they don’t tell you is that the person losing the job is rarely the one equipped to get the new one.

The junior coder whose support role is automated out of existence in Gurugram does not seamlessly transition into the senior machine-learning architect position that opens up in its place. Averages treat human lives like fungible widgets on a spreadsheet, masking a harsh reality: structural displacement. The macro-level comfort that “the economy is fine” means absolutely nothing to the individual who just fell through the gap.

And yet, I distrust the doom as much as the denial, because the prophets of collapse have a poor record. The single loudest prediction that half of all entry-level white-collar jobs would vanish within a few years, has simply not materialised at the level of the whole economy. New kinds of work are already appearing that did not exist five years ago.

Moving Beyond the Dread of AI Changing the Job Market

So what does one honestly tell the young man at the portal? Not you are doomed. That is a lie, and a cruel one. But not relax, it always works out, either. It does not always work out for the specific person.

What one can tell him is, the fact that the jobs have not all gone yet is not proof that he is safe. It is a head start. Very likely the only one he will get. He can spend it refreshing a portal that is not going to refill, or he can spend it moving.

But moving toward what? Not a magic future-proof degree; no one can hand him that. Toward a better question. The video asks will AI take my job? which has no answer and offers only dread. The question he can act on is narrower. Which part of what I do is the routine, rule-bound work a machine reaches first and how do I climb above it, into the judgment, the messy human context, the work that still needs a person?

Fear that freezes him at the portal, refreshing, is useless to him. Fear that makes him look hard at which rung is vanishing, and climb toward a firmer one before it goes, might be the most useful thing he feels. The video is wrong. The comfort is wrong. The truth is standing in the space between them, waiting to be looked at.