Degrees are decorations

When AI can write essays, pass bar exams, and generate production-ready code, what exactly does a degree prove? That you sat in a room for four years? That you survived a filtration system designed before the internet existed? The gap between what credentials signal and what employers actually need has never been wider. This isn't an anti-education argument. It's an anti-credentialism one. Education has value. The question is whether a piece of paper is still the best proxy for it.

The original signal

For decades, the degree served a clear purpose. It told employers three things: you can learn, you can finish something hard, and an institution vouched for you. In a world where information was scarce and verification was expensive, that signal was efficient. Hiring managers couldn't easily test every candidate's abilities, so they used degrees as a shorthand for competence. The system worked well enough when the skills a degree taught were the skills the economy needed. But the economy has shifted faster than curricula can keep up.

The signal is weakening

The numbers tell a story of erosion. The Cleveland Federal Reserve found that the college wage premium, the earnings gap between degree holders and non-degree workers, has declined by about 10 percent since 2000 after more than doubling in the two prior decades. Demand for college-educated workers has plateaued. In 2010, there were 1.2 job postings requiring a degree for every one that didn't. By 2020, that ratio had flipped to 0.6. The Burning Glass Institute reported in 2025 that 52 percent of the class of 2023 were working in jobs that didn't require a degree one year after graduation. They called it the "flipped pyramid": steady demand for experienced workers paired with shrinking opportunities for new graduates. Meanwhile, the World Economic Forum found that AI skills now command a 23 percent wage premium compared to just 8 percent for a bachelor's degree in isolation. The market is speaking. It's just not saying what universities want to hear.

The hypocrisy gap

Companies love to announce they've dropped degree requirements. Google, Apple, IBM, Tesla, they've all made headlines for it. A 2025 survey found that one in four employers planned to eliminate bachelor's degree requirements by year's end. LinkedIn data shows skills-based hiring is expanding globally. But there's a catch. Harvard Business School studied what actually happens after companies drop the requirement. The result: fewer than 1 in 700 hires without a degree actually get through. The policy changes, but the practice barely moves. Hiring managers still filter by school name first. The resume with "NUS" or "Stanford" still lands on top of the pile. This is the structural hypocrisy. Companies publicly champion skills-based hiring while their internal processes still run on credential-based filtering. The degree requirement was never just a checkbox, it was a cultural assumption baked into how organizations evaluate people.

AI changes the calculus

Here's where it gets interesting. AI doesn't just threaten certain jobs. It threatens the entire logic of credentialism. A degree used to prove you could produce certain kinds of output: research papers, code, financial models, legal briefs. Now anyone with a laptop and an AI subscription can produce degree-level output in a fraction of the time. A Gallup survey in 2026 found that 42 percent of bachelor's degree students have reconsidered their major because of AI. They can feel the ground shifting. The "death of the junior developer" thesis crystallizes this. If AI handles the kind of work that junior developers typically do, what exactly does a computer science degree prepare you for? Steve Yegge's widely discussed essay argued that AI is simultaneously eliminating entry-level tasks while raising the bar for what experienced workers need to know. A Dallas Fed paper from early 2026 found exactly this pattern: AI is reducing entry-level hiring while raising wages for experienced workers in the same occupations. The degree was supposed to be the on-ramp. But if AI removes the first few rungs of the ladder, the on-ramp leads nowhere.

The Singapore context

In Singapore, challenging the value of a degree isn't just contrarian, it's almost taboo. The pressure to get a degree is cultural, not just economic. The education system is an arms race from secondary school onward: tuition after school, weekends lost to revision, constant ranking through L1R5 scores and GPA. Everything is a number. Everything is a race. A Lee Kuan Yew School of Public Policy paper examined credential inflation in Singapore's economy, noting how rising educational attainment hasn't necessarily translated to proportionally better outcomes. The Singapore Institute of Technology acknowledged that "a bachelor's degree is no longer sufficient for a career spanning 40 years." Meanwhile, a Milieu Insight study found that 38 percent of Singaporean respondents said a degree is only "moderately important," with alternative paths like skills-based learning gaining ground. The cultural shift is slow. Parents still equate academic success with personal and family success. But the gap between what the system optimizes for (grades, credentials, institutional prestige) and what the economy actually rewards (adaptability, skills, proof of work) keeps widening.

What replaces the signal?

If degrees are losing their signaling power, what fills the void? Three things are emerging. Portfolios and shipping history. The "portfolio over resume" movement is real, even if incomplete. Employers increasingly want to see what you've built, not where you studied. GitHub profiles, published writing, shipped products, these are harder to fake than a transcript. But this shift advantages people who have the time, resources, and awareness to build in public. Skills assessments and certifications. Micro-credentials, AI-specific certifications, and practical assessments are growing. They're faster, cheaper, and more targeted than a four-year degree. LinkedIn's 2025 skills-based hiring report showed that skills-based talent pools are 12 to 25 times larger than traditional credential-based pools, depending on the region. Public work and demonstrated thinking. Blog posts, open-source contributions, conference talks, YouTube tutorials. These don't just show what you know, they show how you think. In a world where AI can generate competent output, the ability to think clearly, make judgment calls, and communicate reasoning becomes the real differentiator.

The counterargument is real

Let's be honest about what degrees genuinely provide. They teach you to think across disciplines. They expose you to ideas you wouldn't encounter otherwise. They build networks that compound over decades. The college experience creates a structured environment for intellectual growth at a formative age. These are real benefits. The question isn't whether they have value, it's whether they're worth $50,000 to $200,000 and four years of opportunity cost, especially when the signaling function that justified much of that investment is eroding. A US News opinion piece argued that a well-designed bachelor's degree develops "critical thinking, ethical decision-making, complex problem-solving, and clear communication," the very skills AI can't replicate. That's a fair point. But it assumes the degree is well-designed, that students actually develop those skills rather than just optimizing for grades, and that there's no cheaper way to acquire them.

The real question

This isn't about whether education matters. Of course it does. Learning to think, to question, to synthesize, these are foundational human capabilities. The question is whether a four-year, six-figure credential is still the best vehicle for that education, or whether it's become a decoration: impressive on the wall, increasingly irrelevant in practice. The answer probably isn't binary. For medicine, law, and engineering, credentialing serves a public safety function. For knowledge work, creative work, and technology, the gap between what a degree signals and what the work demands is growing every year. The students reconsidering their majors because of AI aren't confused. They're paying attention. The companies quietly dropping degree requirements aren't being charitable. They're being rational. The hiring managers who still filter by school name first aren't wrong about wanting a signal. They're just using an outdated one. Degrees aren't worthless. But they're increasingly decorative, a nice thing to have that says less and less about what you can actually do. The sooner we admit that, the sooner we can build something better.

References