The Question
For a century, the formula was simple. Stay in school. Get a degree. Find a job that pays a stable salary. The credential was the signal — proof to an employer that you had the knowledge, the discipline, and the social fitness to do the work. It worked because the skills the degree certified were genuinely scarce and genuinely hard to replicate.
That assumption is now being dismantled at speed. AI systems in 2026 can pass the bar exam, write production-ready code, draft financial analyses, diagnose medical images, and summarise legal contracts with accuracy that rivals specialists. The knowledge the credential was supposed to certify is no longer scarce. What remains scarce — and what the market is beginning to price accordingly — is something harder to test in an exam: the ability to think differently, build trust, navigate ambiguity, and create things that have not existed before.
What the Evidence Shows
The cracks in the credential system were visible before AI accelerated them. Over the past decade, large employers such as IBM, Google, and Apple, along with many US state governments, announced they were dropping degree requirements for many roles. But announcements are not hiring. A 2024 study by Harvard Business School and the Burning Glass Institute, which tracked 11,300 roles at large firms, found that removing the requirement raised the hiring of workers without a degree by only about 3.5 percentage points, and that the promised gains showed up in fewer than 1 in 700 hires.
Now AI is making those cracks structural. When an AI tool can generate a competent first draft of a legal brief in two minutes, the starting lawyer who spent three years memorising case law has a shrinking advantage over the paralegal with sharp judgment and communication skills who knows how to direct and edit that AI output. A 2018 McKinsey Global Institute report found that demand for "higher cognitive skills" — creativity, critical thinking, complex communication — was projected to rise by 19% by 2030, while demand for basic data processing and information collection (tasks degrees often certify) was expected to fall sharply.
"The increased opportunity promised by Skills-Based Hiring has borne out in not even 1 in 700 hires last year."
— Harvard Business School and the Burning Glass Institute, February 2024The wage data shows a degree still pays, though less than it did. A 2026 working paper by economists José Azar, Mireia Giné, and Javier Sanz-Espín, using Census survey data, estimated that US college graduates earned about 78% more per hour than workers without a degree, down from about 87% in 2022 and back to mid-1990s levels. Entry-level graduates are feeling the squeeze most: unemployment among 22- to 27-year-olds with bachelor's degrees was 5.6% in March 2026, according to the New York Fed. Employers do prize creative thinking — the World Economic Forum's 2025 Future of Jobs survey ranked it among the fastest-rising skills — but that has not yet turned into a broad salary premium over degrees.
AI has made knowledge cheap. What it cannot manufacture is a person who sees things differently.
Why This Is Happening
AI commoditises knowledge faster than education systems can respond. A law degree takes three years to complete. The knowledge it certifies is now reproducible by a language model in seconds. Education institutions move on decade-long timescales. The labour market moves on quarterly ones. The gap between what degrees certify and what employers actually need is widening with every AI capability jump.
Employers are discovering that credentials are poor proxies for output. The rapid growth of skills-based hiring platforms — LinkedIn Skills Assessments, Coursera for Business, Multiverse apprenticeships — is giving employers an alternative signal. When you can test whether a candidate can actually do the job, the degree becomes less important. The credential existed partly because there was no better way to evaluate candidates at scale. Now there is.
The work that remains after AI is inherently creative. AI handles volume, consistency, and pattern recognition. The tasks left for humans are the ones AI finds hardest: generating truly novel ideas, reading a room, making ethical judgments, building authentic relationships, and taking responsible creative risks. These are not skills you learn in a lecture hall. They are cultivated through curiosity, experience, and deliberate practice — none of which require a specific credential.
What Could Happen
By 2032, skills-based hiring becomes the dominant paradigm in technology, finance, marketing, and consulting. Degree requirements persist in regulated professions — medicine, law, engineering — but decline everywhere else. Salary data increasingly favours demonstrable creative output over paper qualifications. A generation of workers who built portfolios, shipped products, and cultivated real-world skills without completing traditional degrees out-earns peers with expensive credentials in key knowledge-work sectors.
Rather than replacing credentials, creativity becomes an amplifier of them. Employers still use degrees as a baseline filter — but within that filter, the candidates who command premium salaries are those who combine formal training with demonstrable creative adaptability. The degree stays relevant; it just stops being sufficient on its own. This is likeliest in large corporations and regulated industries, even if skills-based hiring spreads faster among startups.
Public anxiety about AI-generated misinformation and unsafe outputs drives a regulatory and cultural backlash. Employers retreat to credentialed professionals as a trust signal — a way of holding someone accountable within a licensed framework. Governments mandate credentialing for AI-adjacent roles. The degree gets a second life not because it certifies knowledge but because it certifies accountability. Creative talent without formal qualifications finds itself locked out of newly regulated sectors.
What Can We Do
Whether you are entering the workforce, mid-career, or advising someone who is, the shift toward creative skills changes what it makes sense to invest in.
Build a portfolio, not just a CV. In a world where skills matter more than credentials, proof of output matters more than proof of study. Start creating things — writing, design work, code, analysis, video, products — and publish them. A portfolio of real work is more credible to a skills-first employer than a degree certificate from a name institution.
Learn to use AI as a creative partner, not a threat. The workers who will command premiums are not those who resist AI, nor those who outsource everything to it. They are the ones who know how to direct it, push back on its outputs, and add the layer of genuine originality that AI cannot generate. Practice working with AI tools in your field — treat it like learning to use a very powerful instrument.
Invest in the skills AI finds hardest. Persuasion, storytelling, taste, empathy, the ability to hold a room's attention, the judgment to know when a technically correct answer is the wrong one — these are not soft skills. They are the hard frontier of human capability. Pursue them deliberately, whether through workshops, mentorship, performance, or sustained creative practice.
If you are a student, choose your degree strategically. A credential still matters in regulated professions and in large organisations that use them as filters. But what you do alongside the degree — the projects, the internships, the things you ship — may matter more than the degree itself. A mediocre degree from a strong university with a compelling portfolio of real work will increasingly beat an impressive degree with nothing to show for it.
If you are an employer, update your hiring signals. Degree requirements that were never explicitly justified are now actively harmful — they filter out creative talent that would thrive in the role. Defining what the job actually requires, and testing for that directly, produces better hires and a more diverse candidate pool. Skills-based hiring is not a political position. It is a competitive one.
- Burning Glass Institute & Harvard Business School — Skills-Based Hiring: The Long Road from Pronouncements to Practice, 2024
- Harvard Project on Workforce — summary of the skills-based hiring study
- Federal Reserve Bank of New York — The Labor Market for Recent College Graduates
- Bloomberg Law — Unemployment for recent college grads remains high, NY Fed says, 2026
- The College Investor — The college wage premium just fell to its mid-1990s level (Azar, Giné & Sanz-Espín), 2026
- McKinsey Global Institute — Skill shift: Automation and the future of the workforce, 2018