Something Has Felt Off
A graduate scrolls job listings three months after finishing a degree, and the applications have started to blur together. The grades were there. The work was done. Nobody mentioned that the degree wouldn't be enough on its own — that "enough" itself had quietly moved while the system meant to prepare people for it stayed the same.
A parent, elsewhere, watches this happen to a son or daughter, wondering whether the thing they paid for, encouraged, or sacrificed for still does what it used to do. Someone else, a few years into a first job, notices that the role which was supposed to teach them the rest of what they needed to know simply isn't doing that either.
None of this is confined to one person's experience, and none of it is really about individual failing. Something bigger has shifted underneath the whole system, and understanding the shape of it tends to make the discomfort make a lot more sense.
It's Not Just the AI Headlines
Most coverage of AI and universities focuses on cheating — students using ChatGPT to write essays, professors trying to catch them. That's real, and it's growing fast: the share of students admitting to using AI-generated text in graded work has quadrupled in two years. But the cheating story misses the actual problem underneath it.
For as long as anyone can remember, a degree tested whether a student had genuinely learned something — a law student's grasp of statute, a nurse's grasp of physiology, a social worker's grasp of safeguarding law. The exam, the essay, all of it existed to check that the knowledge had genuinely taken root. AI can now produce a fluent, well-structured answer to almost any of those questions in seconds. That doesn't mean cheating became easier. It means the test itself quietly stopped proving what it was built to prove — for every student, not just the ones using the shortcut.
The Job Market Changed Underneath, Too
This part rarely gets connected to the assessment story, and it should be, because it's the same shift showing up somewhere else. Entry-level hiring has fallen sharply over the last two to three years — in some sectors by close to half — and it has landed specifically on the youngest workers trying to get a first foot in the door. For the first time on record, newly graduated young people are, on some measures, more likely to be unemployed than the general population.
There is a genuine live debate about how much of this is AI and how much is the shift to remote working making it harder and more expensive for a business to train someone up close — nobody has fully settled that argument yet, and the uncertainty is worth holding onto rather than reaching for a tidy villain. Whichever mechanism dominates, the effect is the same: the traditional deal — get the degree, then learn the rest on the job — is breaking from both ends at once. The degree does not prove what it used to, and the job that was meant to finish the training is harder to get and, once secured, often does not provide the training either.
When This Deserves Attention
For anyone navigating this right now — studying, applying, hiring, or advising someone who is — the frustration involved has a real structural cause rather than a personal one. Some of the proposed fixes are genuinely promising: universities embedding real workplace supervision directly into degrees rather than leaving it to chance afterwards, and a growing recognition — from inside industry, not only from universities defending themselves — that employers dropped their side of an old bargain and need to pick it back up. Neither side can fix this alone, which is worth remembering before the instinct to blame just one of them takes hold.
Topics: #HigherEducation #AIAndEducation #GraduateEmployment #UniversityAssessment #EntryLevelJobs
Further Reading
These sources dig deeper into the topics covered here:
On assessment and AI:
- HEPI and Kortext's 2026 student survey on generative AI use in higher education
- A 2026 natural-experiment study on assessment validity published via MDPI's special issue on generative AI in higher education
On the graduate job market:
- The World Economic Forum's Future of Jobs Report 2025
- Stanford's 2026 AI Index, tracking entry-level hiring trends by age and sector
How This Essay Reflects YFL Values
This piece follows Information Without Instruction — it names a structural shift affecting students, graduates, and families without telling any individual reader what to do about their own situation. The aim is recognition and understanding first: knowing that a struggle is systemic rather than personal is often the first genuinely useful thing a person going through it can hear.
Related YFL Essays and Resources
AI and Higher Education — the full Repositorium essay — The complete argument behind this piece, with full research and evidence.
In Other Words: Higher Education Has Its Own Debt to Pay from 1999 — The same underlying shift, approached through the sector’s own financial history.
Academia and the Marketplace — The fuller account of what market pressure has already displaced inside UK universities.
Why University, Of All Things — The longer history of why higher education institutions exist at all, and who they have always served.
Poised: Ambition, Risk, and the Nervous System That Gets There First — On the physical experience of standing at a threshold with no guarantee of what's on the other side, the same territory a graduate facing this reckoning is standing in.
HWTK: Why Ambition and Anxiety Feel the Same in Your Body — Same chest, same stomach, same looping thoughts — the bodily experience behind the scrolling-job-listings scene this piece opens with.
In Other Words... The Drop-off at Uni — the other end of the same journey: what four decades of research says about the grief and relief that arrive together when a young person leaves for university in the first place.