Elon Musk thinks we are heading towards a world where “no job is needed”, with AI doing “everything” and humans living on some form of universal basic income and free to do whatever they want. Nvidia’s Jensen Huang, meanwhile, predicts the next generation of millionaires in the medium term will be electricians, plumbers and tradespeople, as AI drives a global boom in data centres and physical infrastructure.
Today’s AI and robotics trends suggest they may both be right.
Robots can now write production-grade code – with frontier AI systems like Claude already matching or outperforming human developers on many standard benchmarks – and large language models pass professional exams such as the US Uniform Bar Exam at roughly human levels. In that fast-changing context, the old assumption that “white-collar work is safe” unravels fast. Even in hospitals, robotic-assisted surgery has gone from pilot to routine practice. Abu Dhabi’s Sheikh Shakhbout Medical City has completed more than 500 robotic surgeries since 2020, while Cleveland Clinic Abu Dhabi performed more than 400 in 2024 alone.
As software starts to eat white-collar tasks and machines take over everything from diagnosis to professional services, it is increasingly plausible that the highest demand – and some of the highest incomes – will sit with people who build, wire, repair and care in the physical world: electricians, construction crews, robotics technicians and caregivers. In that world, where does the traditional university degree fit – especially the classic path into a stable, entry-level office job that appears to be the first thing AI is automating?
To answer that, we should start with evidence, not sentiment. Despite the anxiety, the degree is not dead. In the UK, official labour-market data still shows graduates earning more, enjoying higher employment rates and accessing better-quality jobs than non-graduates. OECD’s Education at a Glance 2025 tells the same story globally. On averages alone, the degree still pays.
But averages conceal the deeper reality. The problem isn’t higher education itself; it’s a university system designed for the industrial age now colliding with a post-industrial labour market. In mature markets, the traditional white-collar entry route is narrowing. The Financial Times recently reported that top UK employers have cut graduate hiring for a second straight year, even as applications for every vacancy have tripled since the early 2000s, now hitting around 140 for every role. The picture is similar in fast-growing economies. Research published in the Journal of Human Resources shows that while lifetime returns remain positive, the early-career premium is shrinking: too many young graduates are competing for too few entry-level office jobs at precisely the moment AI is automating many of those same tasks.
The real problem is that universities are preparing students for the wrong century. Much of the system still behaves as if the goal is to prepare people for 20th-century industrial roles: stable firms, predictable job descriptions, linear careers. Yet the emerging economy looks nothing like this. Routine cognitive work is being devoured by algorithms at a pace few are willing to acknowledge. At the same time, demand is exploding for hybrid skill-sets – part technical, part human; for hands-on trades tied to AI infrastructure; new and growing economic sectors; and for people who can continuously re-skill as technology and geopolitics keep shifting.
For more than a century, universities were built around a simple idea: spend a few years on campus in your early 20s, collect a credential and carry that signal for life. That logic worked when knowledge was scarce, changed slowly and careers were linear. None of that is true anymore.
We are now in the midst of a fundamental shift from the industrial-age university built for a world defined by scarcity of knowledge and stability of work to the AI age that is defined by abundance of knowledge and instability of work. Only the adaptive universities will survive – not the biggest, wealthiest or oldest, but the ones that create real value in a fluid global economy.
What is needed now is a shift from one-off schooling to a lifelong operating system. We need universities that can offer shorter credentials, stackable modules, re-skilling academies and rapid-response courses whenever industries or technologies suddenly shift. Graduates should not leave with a static certificate, but with a “learning passport” that updates as the world does.
This also requires a shift in geography. Forward-looking universities should open themselves to ecosystems: partnering with employers, trades, start-ups, civic bodies and global platforms. Apprenticeships and real-world projects should count as much as exams. Learning that happens in factories, logistics hubs, hospitals, creative studios or community organisations needs to be recognised and credentialed.
So, should young people still go to university? Yes – but not blindly. The old promise that any degree, from anywhere, at any price, guarantees upward mobility is over. What matters now is the alignment between what is learned, how it is taught and the realities of a changing labour market.
In the AI age, the degree is no longer the product. The learner is.
