Imagine joining your first job today. Your manager asks you to prepare a report. You use AI to finish it in minutes. You write code without spending hours debugging it. You analyse data without having to work through spreadsheets line by line.
You finish faster, your company saves time, and everyone seems to win.
But there is a question nobody has fully answered: if AI does all the work through which you were supposed to learn your job, how will you become good enough to do it without AI?
advertisement
This is the uncomfortable question emerging from Microsoft’s latest workforce move, as the technology giant offers voluntary buyouts to around 60,000 eligible employees across India and three other countries.
On October 27, former Microsoft employee Merill Fernando argued that the company needs fewer people and more AI agents to work efficiently. Fernando, who left Microsoft three months ago to start his own company, said layers of management and organisational silos were slowing work down.
His argument was about how Microsoft operates, not a confirmation that AI caused its buyout programme. But it comes at a time when companies are asking how much work smaller teams can accomplish with AI.
And for millions of employees, the question goes beyond whether their current job is safe. It is whether the career they are trying to build will still have a starting point.
THE WORK THAT MAKES YOU AN EXPERT IS THE WORK AI CAN DO
Nobody becomes an expert on their first day at work. A young analyst learns by building spreadsheets, spotting errors and figuring out why the numbers do not add up.
A software engineer learns by writing code, breaking things and spending hours finding the bug. A marketing executive learns by testing campaigns that fail. A journalist learns by making calls, checking documents and discovering that the first version of a story is often wrong.
Much of this work is repetitive. It can also be frustrating. But it is where people develop judgement.
You learn which numbers look suspicious, which questions to ask, when a source is unreliable and why a seemingly correct answer does not make sense. You begin to recognise patterns because you have spent time dealing with them yourself.
Now imagine AI doing much of that work before a beginner gets the chance.
The report is already prepared. The code is already written. The data has already been analysed. The first draft is ready before the new employee has figured out how to begin.
The employee may become more productive. But will they develop the same depth of understanding?
AI can explain a concept, suggest a solution and help someone practise. The risk arises when using AI becomes a substitute for learning how to do the work, rather than a tool that helps people learn it.
EVEN THE PEOPLE BUILDING AI ARE RETHINKING JUNIOR JOBS
This is not just a fear among workers watching technology from the sidelines.
Kevin Sneader, Goldman Sachs’ president of Asia Pacific ex-Japan, said in October that young employees were increasingly starting their careers by managing AI agents.
That is a major change from the traditional career path, in which employees spend years doing analytical and preparatory work before moving into supervisory roles.
Anthropic, the company behind Claude, has also reported a change in its technical workforce. Its October 2026 research found that junior employees accounted for 14% of its technical workforce, down from 21% in early 2025. The share of junior technical hires fell from 27% to 9%.
These figures do not prove that AI alone caused the decline. Hiring priorities and business needs also change. But they offer a glimpse of the dilemma: even companies building AI systems may need to rethink how they bring junior employees into the profession.
The concern is particularly serious because entry-level jobs are not merely a way to earn a first salary. They are where companies develop their future senior engineers, analysts and managers.
STARTUPS WANT SMALLER TEAMS. WHAT HAPPENS TO NEWCOMERS?
The pressure is not limited to technology giants.
Elevation Capital’s report on AI adoption in Indian startups found that 52% of surveyed chief executives were freezing hiring in certain functions or actively reducing team sizes. Engineering, marketing, customer support and operations were among the functions affected.
For a startup operating on a tight budget, the attraction of AI is straightforward. A small team can write code, analyse information, produce content and handle routine customer queries faster than before.
A founder may no longer need to hire as many people to get a product off the ground. That can help a business survive and grow.
But multiply that decision across hundreds of companies, and the consequences for fresh graduates become harder to ignore. Even without mass layoffs, fewer entry-level vacancies can mean fewer chances to get that first job.
And if fewer people enter a profession today, the shortage may show up years later, when companies need experienced people who understand the work beyond what an AI system can produce.
WHO WILL TEACH THE NEXT GENERATION?
This is where the debate about AI and employment becomes bigger than job cuts.
Companies cannot expect every new recruit to arrive with years of experience. Nor can they expect future managers to emerge from nowhere. Expertise is built through practice, feedback, mistakes and gradually taking responsibility for more difficult decisions.
If AI handles routine tasks, companies will have to find new ways to give beginners that experience. Junior employees might review AI-generated work, investigate exceptions, work directly with customers or tackle problems that do not have easy answers.
Managers will need to teach them why a solution works, not just ask them to produce it faster.
Workers, too, will need to learn when to use AI and when to work through a problem themselves. Knowing how to generate an answer is not the same as knowing whether it is correct.
The danger is assuming that because AI can perform a task, humans no longer need to learn how that task works.
Five or ten years from now, companies will still need people who can make difficult decisions, recognise mistakes and take responsibility when things go wrong. The question is whether those people will get enough opportunities to develop those skills today.
Microsoft’s buyouts are the immediate headline. The bigger concern is what happens to the career ladder if companies automate the very work that once helped people climb it.
AI may change how we learn our jobs. Employers now have to make sure it does not eliminate the chance to learn them at all.
– Ends