Ten years is a long time in engineering. The student who selects a specialty today will be graduating around 2030, will spend a couple of years getting their first experience, and will be starting to hit their stride somewhere around 2034-2036. The specialty they select today defines what they know, whom they know, and what type of problems they are good at solving over this entire period.
The problem is that nobody knows what 2036 looks like exactly. Predicting technology is a good way to be wrong. The only thing one can predict reliably is which fields are likely to have enough underlying momentum that their basic direction is unlikely to turn around, and what types of engineering training will still be useful regardless of the specific technological content.
The discussions about branch selection in India seem to fall within two bad paradigms.
The first paradigm is following the current hype. A few years ago, every student in India seemed to want Computer Science as it provided the most opportunities. In the years before that, IT was the craze. Now the attention has shifted to AI and Data Science as those are the big corporate buzzwords. Staking a four-year commitment on whatever is hot at the moment is a dangerous gamble, because market hype describes the present state, while supply and demand of a particular skill always tends to equal out over time. All the students now stampeding into AI programmes in 2025 will be graduating in 2029, and all will flood the job market at the same time.
The second bad paradigm is opting on the basis of what parents or relatives deem secure. Professional security is never a matter of a particular subject. Professional security lies with the ability and usefulness of the student, and only then the subject itself. A well-taught problem-solving oriented Mechanical Engineering graduate will find a good career in 2035. An unfocused Computer Science student who "drove through" and managed to clear the papers will find life difficult, no matter how sought-after the branch seemed while admissions were open.
Some things can be said with reasonable confidence about where engineering is heading. Energy is not a passing trend. The scale of infrastructure required to decarbonise the global economy, the solar installations, the grid upgrades, the battery storage systems, the hydrogen production facilities, represents decades of engineering work. Electrical engineering, mechanical engineering, and chemical engineering all have direct relevance to this. It is not a niche. It is one of the largest coordinated engineering efforts in human history.
Semiconductors have become a strategic national priority in India in a way that wasn't true five years ago. The government investment in domestic chip manufacturing is not a budget line that disappears in the next election cycle. It's a multi-decade infrastructure commitment. Electronics and VLSI engineering will be relevant in 2036. The question is whether enough students choose it.
AI and machine learning are genuine and lasting. The tools will change. The need for people who understand the underlying principles, who can build, evaluate, and improve intelligent systems, will not disappear. But this is a field where depth matters enormously. A student who studies AI and Data Science seriously, who actually understands statistics, linear algebra, and how models behave when things go wrong will eventually be positioned well.
Civil and infrastructure engineering is less discussed in branch selection conversations than it should be. India's urban infrastructure is under significant strain and is going to require sustained engineering attention for the next several decades. Smart city development, metro expansion, water management systems, coastal infrastructure under climate pressure: none of this is going away.
What the branch selection conversation is really about, even when it doesn't say so explicitly, is risk. A student and their family want to know that the four years and the money and the effort will produce a career worth having.
The honest answer is that no branch guarantees that. What comes closer to guaranteeing it is choosing something the student is genuinely interested in and then engaging with it seriously for four years. Interest sustains attention. Attention produces depth. Depth is what makes someone valuable in a professional context, regardless of what the branch is called.
A student genuinely drawn to how electrical systems work, who finds circuits and power systems and signal processing interesting, will develop more useful capability in Electrical Engineering than in Computer Science they chose because someone said it was safer. In 2035, they'll be working on problems that draw directly on what they actually spent four years paying attention to. The branches worth considering for B. Tech, AI and Data Science, Cybersecurity, Electronics, Electrical, Civil, Mechanical, Computer Science, are all reasonable choices in terms of structural relevance. None of them are going to be irrelevant in ten years. The variable is not the branch. It's what the student does inside it.
Before finalising a branch, find three people who work in that field right now, not faculty, not career counsellors, but working engineers five to ten years into a career in the area. Ask them what they actually do on a Tuesday afternoon. Ask what they wish they'd known before studying what they studied. Ask whether they'd make the same choice now.
Three conversations won't give you a complete picture. They'll give you something better than a ranking or a salary projection: a sense of whether the work itself is something you can see yourself doing. That's the variable that matters most across ten years.