Opinion | Will AI Take Your Job, or Change It?

Every large machine in history scared the people who worked beside it. The loom scared weavers. The tractor scared farmhands. The computer scared clerks. Now it is the turn of software that can write, translate, summarise and code, and the fear feels sharper because these tools touch work that was supposed to be safe, the kind done at a desk with a degree.

An honest answer starts with “we don’t know”: Anyone who tells you with confidence exactly how many jobs will vanish is guessing. Forecasts on this subject have been wrong in both directions for decades. Some predicted mass unemployment from computers and got the opposite. Others said office work was safe and then watched whole departments shrink.

What we can say is more modest and more useful. Technology tends to remove tasks, not whole jobs. A bank clerk lost the task of counting notes by hand, but the bank still needed the clerk for advice, checking and customer care. The job changed shape. Sometimes it shrank. Sometimes it grew. Rarely did it simply disappear overnight.

Tasks, not titles: A helpful way to think about it is to break any job into its tasks. A journalist researches, interviews, writes, edits, checks facts and decides what matters. A new tool may speed up some of these tasks, such as transcribing an interview or summarising a long report. It does much less to help with others, such as knowing which question to ask a nervous official or judging whether a source is telling the truth.

When we look at jobs this way, the question changes. It stops being “will my job exist?” and becomes “which parts of my work are routine, and which parts need judgement, trust and presence?” The routine parts are the ones most exposed. The parts that depend on human judgement tend to become more valuable, not less, because they are what remains scarce.

Who gets the ladder and who gets the trapdoor: The same tool can lift one worker and trap another. A software engineer in a big city has training, savings, a network and time to learn new methods. A data entry operator in a small town has few of those. If a tool does half of the operator’s work, the employer may simply need half as many operators.

This is why the debate cannot be left to technologists alone. Whether AI becomes a ladder or a trapdoor depends less on the technology and more on who has access to training, who owns the tools, and who captures the savings. If the gains go only to firms and the costs fall only on workers, the result will be anger, not progress.

There is another quiet risk. Entry level jobs have always been where people learn. A junior copywriter learns by writing bad drafts and getting them corrected. A junior analyst learns by building simple reports. If tools can do those starter tasks cheaply, firms may stop hiring juniors.

That creates a strange problem. A profession needs experienced people, and experienced people come from beginners who were given a chance. If we remove the bottom rungs of the ladder, we may find in ten years that nobody is ready to take the senior roles. Employers who think about this early can protect their own future by deliberately keeping and training beginners, even when a tool could do the work faster.

Every business that adopts these tools faces a moral choice, whether it says so or not. If a tool saves thirty percent of a worker’s time, there are two broad options. The firm can cut thirty percent of its staff and pocket the saving. Or it can use that time to do better work, serve more customers or build something new.

Neither option is automatically right. Some firms genuinely need to shrink to survive. But it is worth asking who made the decision, who was consulted and whether workers were given a fair chance to learn the new tools before being judged by them. Firms that treat workers as partners in the change tend to keep both trust and skill. Those that treat them as costs tend to lose both.

Governments have three useful jobs here. The first is training. Short, practical courses that help people use new tools, delivered close to where they live and at low cost, matter more than grand national strategies. The second is support for people in transition. A worker who loses a job needs income, time and help finding a new path, not only advice to reskill. The third is clear rules. People should know when an automated system is making a decision about their loan, job application or benefits, and they should have a way to challenge it.

For an individual worker, the advice is not complicated, though it is not easy. Learn to use the tools in your field, so you are the person who knows how to use them well and not the person replaced by someone who does. Build the skills that machines handle poorly: judgement, honesty, clear speech, empathy and the ability to earn trust. Keep a record of your real results, since proof of what you can do will count for more than the name of a course.

Technology does not decide what kind of future we get. People do, through the rules they write, the training they fund and the way they treat each other at work. The real question is not whether machines will become more capable. They will. The question is whether we will use that capability to give more people better work, or to give fewer people more power.

Writer is a columnist and can be reached at mdmuslimbhat@gmail.com