What Are You Actually Preparing For? A guide to choosing a college path in the age of AI*

Not easy choice.

There is a formula that a generation of parents absorbed so deeply it became common sense: go to college, choose a respectable major, get a stable white-collar job, climb the ladder. It was not bad advice for its time. For decades, a university credential was roughly the most reliable bet an ambitious young person could make.

That bet is now much more complicated, and pretending otherwise doesn’t help  young people.

The Scale of What Is Happening

It is worth taking a look at the numbers before drawing conclusions from them. The World Economic Forum's Future of Jobs Report 2025, drawing on surveys of more than 1,000 leading employers across 55 economies, projects that by 2030, 170 million new roles will be created worldwide while 92 million existing jobs are displaced — a net gain of 78 million positions, but a structural churn affecting roughly 22% of the global formal workforce. The IMF estimates that roughly 40% of jobs globally face meaningful exposure to AI, a figure that rises to approximately 60% in advanced economies. Goldman Sachs has estimated that generative AI could expose tasks equivalent to 300 million full-time jobs to some degree of automation.

These are not science-fiction projections. They are the working forecasts that governments and corporations normally use to make decisions. In March 2026 alone, AI was cited as the reason for 15,341 US layoffs — 25% of all job cuts that month, and the single leading cause.

What this means for a seventeen-year-old choosing a college path is not obvious. Anyone who tells you it is probably hasn't thought hard enough about it. But there are things that can be said with reasonable confidence.

The Irony at the Center of This Story

The professions most exposed to AI disruption are not factory jobs. Office and administrative support has the highest task-automation share in the US at 46%, followed by legal at 44% and architecture and engineering at 37%. The WEF identifies bookkeepers, payroll clerks, insurance claims processors, and bank tellers as among the most vulnerable — because their work involves structured, repetitive data tasks that current AI executes efficiently and cheaply.

In other words: many of the white-collar careers that previous generations pursued specifically for safety are now among the most structurally exposed. An AI model can draft a legal memo, reconcile accounts, generate a financial model, and produce the first draft of almost any document a knowledge worker might spend a day on.

What it cannot do — at least not yet, and not reliably — is rewire a hospital operating suite, diagnose a failing HVAC system during a winter emergency, troubleshoot a data center at 2 a.m. when thousands of businesses depend on it, or provide the kind of present, embodied human care that a frightened patient or a grieving family actually needs.

This is why the skilled trades are experiencing something that would have seemed improbable a decade ago: a genuine renaissance in both economic value and social recognition. Electrician employment is projected to grow 9.5% through 2034 — more than triple the average for all occupations. HVAC technician roles are expected to grow 8.1% over the same period. These are not marginal gains. The share of teenagers considering vocational or trade school has more than doubled in six years, from 12% in 2018 to 30% in 2024, and nearly one in four Gen Zers has seriously considered or is actively pursuing a career in the trades.

Part of what is driving this is a detail that tends to get lost in abstract debates about AI and the future of work. The electrician wiring a data center today is building the backbone of the AI boom, where a single error can cost millions in downtime. Ford's CEO Jim Farley has said plainly that America's AI ambitions are running into a labor wall: "I think the intent is there, but there's nothing to backfill the ambition." The data centers that run AI need to be built, cooled, wired, and maintained by human beings with physical skills, and right now there are not nearly enough of them.

None of this means college is obsolete or that the trades are universally superior. It means the old hierarchy — four-year degree respectable, trades somehow lesser — was always more about class signaling than economic reality, and AI is making that clearer faster than anyone expected.

The Question Nobody Actually Asks

Most conversations about college majors skip the most important question. It is not "what pays well?" and it is not "what do I love?" — though both matter. It is this:

What kind of person am I becoming, and will what I am building in myself remain worth something in a world where machines increasingly handle routine intellectual work?

That reframes everything. It asks you to think not about what you want to do in your first job, but about what you are actually developing — what you will be capable of — by the time you finish.

How to Think It Through

Will this still create real economic value?

The first thing to reckon with honestly is whether the field you are considering is growing or shrinking, and whether the specific work you would do within it is the kind machines handle well or handle poorly. There is a meaningful difference between a lawyer who argues in court, reads people, and makes strategic judgments under pressure, and a paralegal who processes standardized documents. AI is transforming the second role far faster than the first. The same pattern runs through almost every profession.

The WEF projects that technological skills will grow more rapidly than any other category through 2030 — AI and big data at the top, followed by cybersecurity and technological literacy. But creative thinking, resilience, analytical thinking, and leadership round out the top ten rising skills. The list is not purely technical. What it describes is a labor market that increasingly rewards people who can do things machines cannot, and those things turn out to span technical, relational, and creative territory.

Students entering highly automatable fields are not necessarily making a mistake — but they need to aim above the routine level of practice, toward the kind of judgment, creativity, or human relationship that actually requires a person to be there.

Does the daily reality of this work suit who you are?

A career you find meaningless is not just unpleasant. It is economically risky. Disengaged people do worse work, adapt more slowly, and burn out in environments that increasingly demand initiative, creativity, and continuous learning. Prestige fades faster than most eighteen-year-olds expect, and salary alone has a well-documented ceiling as a source of motivation.

The useful question is not "would I enjoy this?" — too vague, too present-tense. It is: do I respect the people who do this work? Do I find the actual problems of this field genuinely interesting, not just the idea of the field? Would I care about doing it well even if nobody particularly admired me for it?

Are you building the ability to keep learning, or just acquiring a credential?

The WEF projects that 39% of the key skills required in the job market will change by 2030. That is a remarkable pace of change — it means that a significant portion of what is considered essential expertise today will be obsolete or substantially transformed within the decade. The most durable investment you can make is therefore not in a specific body of knowledge, but in the capacity to learn, unlearn, and relearn under pressure.

This has real implications for how to evaluate a college or a program. An education that teaches you to think rigorously across problems — to analyze, communicate clearly, reason about tradeoffs, and engage confidently with evidence — will serve you across multiple careers in ways that narrowly vocational training may not. The best education is not primarily preparation for a job. It is the development of judgment.

Who will your work actually serve?

This question tends to get skipped because it sounds idealistic. It is in fact one of the more practically reliable filters available. Professions tied to irreducible human needs — healthcare, education, infrastructure, mental health, food, public safety, eldercare — tend to be more durable across economic disruption than professions whose value depends on a particular configuration of technology or market structure that can shift. The WEF's largest growing jobs by absolute numbers include farmworkers, nursing professionals, social workers, and building construction workers. These are not glamorous headlines. But they are evidence that genuine usefulness has a way of remaining economically rewarded even when other things change dramatically.

What kind of life does this path actually produce?

A profession shapes far more than your job title. It shapes your schedule, your stress, where you can live, how much financial volatility you can absorb, what your family life looks like in practice, and over time — whether you realize it or not — who you become as a person. A surgeon, a freelance contractor, a teacher, a tradesperson who owns their own business: these are different jobs, but more than that, they are different lives with genuinely different textures, freedoms, and burdens. That deserves real thought before committing to a path rather than after spending four years and considerable money on one.

Which Paths Look More Durable?

No prediction here is fully reliable. The pace of AI development makes confident forecasting really difficult. But certain patterns are visible.

Fields that combine physical presence, complex judgment, and irreducible human relationship tend to be structurally more resilient: healthcare, nursing, mental health, education, complex engineering, cybersecurity, AI systems management, and skilled trades — particularly electrical work, HVAC, advanced manufacturing, and infrastructure, where the shortage of qualified workers is already severe and growing more so.

Fields organized around routine cognitive processing face the most serious medium-term pressure: standard legal and financial work, basic programming, administrative coordination, data entry, and undifferentiated content production. This does not mean those fields will disappear. It means that surviving and thriving in them will require operating at levels of judgment, creativity, and leadership that AI cannot yet replicate — which raises the entry bar considerably.

The most interesting future profiles are genuinely hybrid: people who combine real technical competence with strong communication, organizational understanding, and the kind of human leadership others actually want to follow. That combination is rarer than it sounds, and rarer than it used to be.

The Most Dangerous Choice Is Not the Wrong Major

It is the absence of any real choice at all — the drift.

The students who end up most exposed are not usually those who chose a difficult or uncertain field. They are those who accumulated debt without a strategy, who chose based on vague prestige without investigating what the actual daily work involves, who let parents or trends or inertia make the decision for them, and who avoided the harder questions because those questions felt too large to confront at eighteen.

That is understandable, but also expensive to sustain.

What Schools and Parents Need to Hear

Parents who are still equating success primarily with prestige — a recognizable university name, a white-collar job title, graduate school as a universal solution — are working with a map that no longer matches the territory. The old status hierarchy was never a complete picture of what a good professional life could look like, and it is now a substantially less reliable guide than it was even ten years ago.

Universities that genuinely serve their students should provide honest, specific information about where graduates actually work and what they actually earn — not marketing language. They should expose students to real work early and often, build technological fluency across every discipline, and teach the kind of adaptive thinking and ethical reasoning that holds up across careers rather than just across semesters.

The Deeper Question

Behind every choice about what to study is a question that the current moment makes difficult to avoid:

What is education actually for?

If it is primarily workforce training — credential acquisition calibrated to current employer preferences — then students will spend their professional lives chasing a moving target, perpetually trying to match skills to a market that AI is restructuring faster than any curriculum can track.

But if education cultivates the things that make a person genuinely hard to replace — the ability to read a situation and make a wise call, to earn trust, to create something that wouldn't have existed otherwise, to lead people through uncertainty, to care competently and honestly about the people they serve — then it confers something durable that no labor market forecast can threaten.

The WEF lists curiosity and lifelong learning, creative thinking, resilience, and leadership among the fastest-rising skills through 2030 — alongside AI literacy and data analysis. The machine is not replacing the curious, adaptable, ethically serious, genuinely skilled person. It is making that person rarer and, in the process, considerably more valuable.

Choose a path not merely to survive the economy coming into view. Choose it to build a life that is intellectually alive, economically grounded, useful to others, and worth the years you will spend living it. That is a harder standard than prestige. It also tends to be the only one that holds up.

What Students Should Optimize For in the AI Era

Beyond salary or prestige — a framework for weighing what actually matters

Dimension	Relative Priority
Adaptability & lifelong learning	Essential — highest priority
Human-centered capabilities	Essential — highest priority
Economic resilience	Essential — highest priority
Purpose & fulfillment	High priority
Service to society	High priority
Technological fluency	High priority
Prestige alone	Low weight — declining signal value

 *Text edited with AI assistance.