For twenty-five years, I’ve sat across the table from teenagers and twenty-somethings at Career Counseling Connecticut, and I’ve watched the anxious question in the room change shape. It used to be “what should I major in?” Then it became “will I get into a good school?” Today, more often than not, it’s something rawer: “Is there even going to be a job waiting for me?”
I want to answer that question honestly, because honesty is the whole job here. And the honest answer is: yes, there will be a job waiting for you. But the door into that job has gotten narrower, the path to it looks different than it did even three years ago, and the young people who understand why it’s narrower are going to walk through it with far more confidence than the ones who are simply hoping AI is a passing headline.
Let me walk you through what the data actually shows, what it doesn’t show, and — most importantly — what to do about it.
The Real Story Isn’t Mass Layoffs. It’s a Closing Door at the Bottom Rung.
If you’ve been half-listening to the news, you’ve probably absorbed a vague sense of dread: “AI is taking jobs.” That framing is mostly wrong, and it’s actually more useful to understand why it’s wrong.
The most rigorous research on this — a Stanford Digital Economy Lab study using payroll records from more than twenty-six million American workers — found something more precise and, frankly, more interesting than a wave of layoffs. Overall employment in the most AI-exposed occupations actually rose. Experienced workers in those fields aren’t getting shown the door. What’s shrinking is hiring — specifically, hiring of workers in their early twenties into entry-level roles in AI-exposed fields, which has been running meaningfully behind their less-exposed peers.
Here’s the mechanism, and it’s worth sitting with for a moment. A company doesn’t need to fire its senior analyst to change its workforce. It keeps her, hands her a tool that drafts the first pass of a report or checks the formulas, and then quietly decides the next junior position isn’t urgent this year. No headline. No layoff. Just a door that opens a little less often. Multiply that decision across enough companies and you get exactly what Handshake’s 2026 graduate report found: job postings for early-career workers are down double digits from pre-pandemic levels, and nearly two-thirds of graduating seniors say they feel pessimistic about their prospects.
I want to be careful here, because catastrophizing helps no one and it’s not who I am. The researchers themselves are explicit that this is a warning light, not a verdict. Aggregate employment is still growing. Companies have real choices about how they redesign junior roles — some are already restructuring entry-level work around verification, client contact, and supervised judgment rather than eliminating it outright. This is a genuine inflection point, not a countdown clock. But families who walk into Career Counseling Connecticut deserve the unvarnished version, because pretending this isn’t happening does a disservice to every young person trying to plan five years out.
Who’s Actually Vulnerable — And Who Isn’t
The pattern in the data is consistent enough now that I can say it with real confidence: the jobs most exposed are the ones built around routine, text-heavy, rules-based tasks — the first-draft memo, the entry-level code review, the repetitive data pull. Those are precisely the tasks that used to be how a twenty-two-year-old earned their stripes. That’s the uncomfortable part. It’s not that young workers are less capable; it’s that the traditional on-ramp — do the grunt work, learn the business, get promoted — is the exact terrain AI is best at covering.
On the other side of the ledger, a fairly consistent list of fields keeps showing up as resilient across every serious study I’ve reviewed: healthcare, skilled trades, education, cybersecurity and compliance, and roles built around hands-on physical work, licensing, or high-stakes human trust. These aren’t “safe because AI hasn’t gotten there yet” — they’re safe because they depend on something AI structurally can’t do: bear accountability, read a room, adapt to an unrepeatable situation in real time, or put hands on a person or a machine.
I’ll also say this, because I think it gets lost in the doom-and-gloom coverage: this is not an argument against technical fields. Software engineering, data analytics, AI/ML engineering, and cloud computing are still among the highest-growth, highest-paying paths available to young people — but the profile of who thrives in them has shifted. The junior developer who only writes routine code is exposed. The one who can architect a system, exercise judgment about what the AI got wrong, and manage a project is not just safe — they’re increasingly the only kind of junior person a company wants to hire, because they can supervise the AI instead of competing with it.
The Skill Stack That Actually Future-Proofs a Career
If there’s one piece of advice I give every student and every parent who sits down with me at Career Counseling Connecticut, it’s this: stop thinking about your career as a major or a job title, and start thinking about it as a skill stack. The young people who are going to do well in the next decade are pairing two things — genuine fluency with AI tools, and the specifically human capabilities that AI can’t fake.
That second category keeps showing up in every credible study I’ve pulled together for this article, almost word for word: critical thinking, emotional intelligence, complex problem-solving, persuasive communication, ethical judgment, and adaptability. Not because these are nice-to-have “soft skills,” but because they are now the load-bearing skills. A software engineer who thinks clearly, writes and speaks well, and can lead a team through an ambiguous problem isn’t competing with AI — they’re the person directing it.
There’s a warning buried in this research, too, and I think it’s one of the more important things for parents to hear. Several 2026 studies — including work out of MIT — found that heavy, uncritical reliance on AI tools actually degrades the very skills that make a person valuable: original thinking, deep recall, a sense of ownership over your own work. In other words, the students who lean on AI to skip the thinking rather than sharpen it are quietly undermining the exact capability the job market is starting to pay a premium for. I tell families this often: learn to use these tools well — that’s non-negotiable now — but don’t let them do your thinking for you. Your judgment is the product. Protect it like one.
What This Actually Means for Your Son or Daughter
Here’s where I get practical, because that’s the work I do every day.
Build proof, not just credentials. A transcript used to be enough of a signal. It isn’t anymore — employers are increasingly stripping degree requirements from job postings in favor of demonstrated ability. That means internships, portfolios, small independent projects, and real experience matter more than they did even five years ago. This is exactly the territory Career Counseling Connecticut works in with students: helping them build something they can point to, not just a list of courses they sat through.
Choose fields with a durable human core. This doesn’t mean abandon technology — it means go into technical fields with your eyes open about what will make you indispensable inside them. Cybersecurity, healthcare-adjacent technology, applied AI work that requires domain judgment, skilled trades with a technology layer — these combine growth with durability.
Get fluent with AI tools deliberately, not passively. There is now a meaningful and growing gap between students who can use AI as a genuine thinking partner — someone who verifies, questions, and directs it — and students who use it as a crutch. Employers can tell the difference, and increasingly, so can the work itself.
Don’t panic about the major. Panic about the muscle. I’ve had families come in convinced that only computer science or engineering is “safe” now. I push back on that, gently but firmly. The data doesn’t support narrowing to a handful of majors — it supports building critical thinking, communication, and judgment no matter what you study, and then applying those muscles to a field you actually care about. A history major who can write, reason, and use AI tools with discernment is in a stronger position than a student in any major who can’t.
Start the conversation earlier than you think you need to. This is the single biggest shift I’ve seen in twenty-five years of doing this work. Career planning used to start junior year of high school. Now, given how fast the landscape is moving, the families who are best positioned are the ones having these conversations two or three years earlier — not to force a decision, but to build the exploratory experiences and the honest self-knowledge that make a good decision possible later.
The Part I Want You to Hold Onto
I left a career as a prosecutor and an SEC enforcement attorney because I believed helping young people find real clarity about their own path mattered more to me than anything I could do inside a courtroom. Twenty-five years later, I believe that more than ever — and I believe it especially now, in a moment when the noise around AI can make a genuinely promising future feel like a closing door.
It isn’t closing. It’s changing shape. The young people who walk through it are going to be the ones who built real judgment, who can work alongside these tools instead of being replaced by the version of themselves that only did rote, repeatable work — and who had someone in their corner helping them see the difference early enough to act on it.
That’s the work we do, every single day, at Career Counseling Connecticut. If your family is trying to make sense of what all of this means for your particular kid, with their particular strengths — that’s exactly the conversation I want to have with you.
Let’s build the next chapter of your child’s, with clear eyes and real confidence.
— Daryl Capuano, Career Counseling Connecticut