Let's be honest about the question. When AI can pass the bar exam in the 90th percentile, write a medical school thesis, code a production application, and summarize any textbook in minutes — it is completely reasonable to look at four years and $150,000 and ask: what exactly are we paying for?
That question deserves a serious answer. Not a defensive one from someone protecting an institution. Not a dismissive one from a tech contrarian. A real one — grounded in what's actually happening in the labor market, in education, and in the lives of people navigating this transition.
Here's what the data says. And here's what it means for you.
The Ground Is Shifting — Faster Than Anyone Expected
Thirty-three percent of companies have already dropped bachelor's degree requirements for at least some roles. Google, Apple, IBM, Accenture, and Deloitte are among the names on that list. ZipRecruiter data shows job postings requiring a degree fell 10% between 2022 and 2023 alone. Seven in ten hiring managers now say relevant experience matters more than a diploma when they're making final decisions.
This is real. It's happening now. It's not a prediction.
And the AI disruption to actual jobs is equally real. Goldman Sachs estimated 300 million jobs could be affected by automation. The World Economic Forum projects 92 million roles will be displaced by 2030. Early-career software developers — people who did everything right, got the degree, got the entry-level job — saw a nearly 20% employment decline between 2022 and today as AI ate the bottom of the coding stack.
So the skepticism about traditional education is not irrational. The people asking "what's the point?" are not naive. They're pattern-matching correctly on a world that is genuinely changing.
But Here Is the Part That Gets Left Out
The same World Economic Forum report that projects 92 million displaced jobs also projects 170 million new jobs created — a net gain of 78 million roles. The fastest-growing categories: AI and machine learning engineers, big data specialists, cybersecurity analysts. Growing at 80–140% rates.
Bachelor's degree holders still earn a median of $80,000 per year. High school graduates earn $47,000. That $33,000 annual gap is near an all-time high — not shrinking. Federal Reserve research puts the ROI of a college degree at 12.5% annually, beating the long-term return of the stock market. A 2025 bachelor's graduate can expect lifetime earnings of $7.95 million.
Harvard Business School researchers found something important about the companies that "dropped degree requirements": in practice, fewer than 1 in 700 applicants without a college degree actually got hired. The policy changed. The hiring behavior barely did.
This is the part of the story that gets lost when the headline is "Google drops degree requirement." The gap between what companies say and what actually happens in recruiting is enormous.
What AI Actually Threatens — And What It Doesn't
Here's the uncomfortable truth that neither side of this debate wants to say clearly: AI is most dangerous to the people who only learned how to execute tasks, not how to think.
The lawyers who will lose their jobs are the ones who drafted the same contract templates 500 times. The developers who are struggling are the ones whose entire value was writing boilerplate code. The analysts being replaced are the ones who ran the same Excel model on repeat.
AI does not threaten the person who understands why those contracts matter, who can architect a system rather than implement it, who can interpret what the model output means and make a judgment call about what to do next.
A 2025 Microsoft study found that heavier reliance on AI correlated with lower critical thinking. That finding should stop you cold. It means the people most at risk in the AI era are not the ones without degrees — they're the ones, with or without degrees, who outsource their thinking. And the skills that protect you — analytical reasoning, ethical judgment, complex problem-solving, clear communication under ambiguity — are exactly what a rigorous education, done right, is supposed to develop.
The degree is not what protects you. The thinking is. The degree was supposed to be where you learned to think.
So Should You Go to School?
The honest answer is: it depends on what you're trying to do — and it depends enormously on the major, the institution, and how you use the experience.
Finance majors generate a lifetime ROI of 1,842% on their educational investment. Computer science: 1,752%. Some humanities majors at expensive private colleges: negative ROI for 25% of graduates.
A degree from a strong institution in a high-demand field, completed in four years with real internships, real projects, and a real professional network? Still one of the best investments a young person can make.
A degree in a saturated field from a school with no industry connections, finished in six years with $180,000 in debt? The math is brutal.
The question was never really "should I go to school?" It was always "what am I actually getting, and is it worth what I'm paying?" AI just makes that question impossible to avoid any longer.
The Third Path That Most People Miss
The real story in the enrollment data is not dropouts and not traditional degrees. It's certificates.
Undergraduate certificate enrollments grew 6.6% in fall 2025 — the fastest-growing credential category. The people flooding into certificate programs are not naive. They're reading the room. They understand that a 12-week credential in AI tools, a Google data analytics certificate, a cloud certification from AWS — these are not consolation prizes. In a skills-first hiring environment, they are increasingly what gets you in the door.
Lucy Guo dropped out of Carnegie Mellon. She went on to co-found Scale AI, which was acquired by Meta for $29 billion, making her the youngest self-made female billionaire in Forbes history as of June 2025. She is a genuine inspiration. She is also one person in a sea of college dropouts who did not become billionaires — because the selection effects around who successfully drops out to build a company are extreme.
The realistic version of the third path is not "drop out and build a startup." It's: pursue the knowledge and credentials that are most directly connected to what you want to do, by the fastest and most cost-effective route available. Sometimes that's a four-year degree. Sometimes it's a two-year program. Sometimes it's a year of intensive self-study, certifications, and portfolio projects.
What Actually Matters Now
This is the part worth writing down.
The WEF asked thousands of employers what they actually need from workers in 2025 and beyond. The top answers were not "degrees." They were: analytical thinking. Resilience and adaptability. Creative thinking. The ability to keep learning.
Think about what that list really says. It says the single most important thing you can do — with a degree or without one — is become someone who does not stop learning when the formal education ends. Someone who sees AI as a tool to amplify their thinking rather than a replacement for it. Someone who can figure out what matters in a situation, make a judgment call, and communicate it clearly.
Sam Altman, the CEO of OpenAI, said he's "envious of Gen Z college dropouts" who have the time and mental space to build. He also said he believes that by 2035, today's graduates will be leaving to explore opportunities we can barely imagine yet. Both things can be true simultaneously: the old routes are less automatic, and the new landscape has more room than the doom-scrollers suggest.
The Question Underneath the Question
When people ask "does school even matter anymore?" they're often really asking something else: Am I going to be okay? Is the path I'm on still going to lead somewhere? Can I build something real without following the route I was told to follow?
The answer to those questions is yes. Emphatically. But it requires being honest about what you're building and why — not just collecting credentials and hoping they age well.
The people who will thrive in the AI era are not the ones with the most impressive diplomas on their wall. They're the ones who never stopped asking real questions, who treat every tool — including AI — as something to master rather than something to be managed by, and who understand that the most future-proof thing they can do is become genuinely, deeply good at something that matters.
School can still be part of that. So can everything else.
What you cannot afford is to stop learning the day you get the piece of paper — or the day you decide not to.
