News Story

Breadth vs Depth in an AI world: Rethinking graduate skills for 2036

Monday 30 Mar 2026

Last week, KMi’s Prof John Domingue joined an influential panel at Keele University to discuss one of the most pressing questions facing higher education: what kind of graduates will society need in an AI‑shaped future? The event brought together leaders from education, industry and policy, including former Home Secretary Charles Clarke and Times Higher Education Editor Chris Havergal, to reflect on how universities should rethink learning for the workforce of 2036.

In his contribution, Prof Domingue challenged the “generalist versus specialist” framing. The real divide emerging today, he argued, is between “knowledge athletes”, graduates trained to recall and produce information, and “AI‑era operators” who can direct, verify and govern intelligent systems. As AI accelerates, the value of routine cognitive tasks declines; what matters instead is the ability to steer complex, AI‑augmented workflows safely and productively.

Using vivid analogies, he described the shift from training knowledge “engines” to training “F1 drivers”: graduates who can set goals, manage constraints, use advanced tools intelligently and maintain oversight of automated systems. Looking ahead, many professionals won’t simply complete tasks themselves, they will design “knowledge factories”, orchestrating collections of agents, tools and checks that generate repeatable, high‑quality outputs at scale.

Prof Domingue emphasised that inclusion must be at the centre of this transition. If only a privileged minority learn to “drive” AI systems, he warned, others risk being left as passengers in an economy increasingly powered by automation. Universities therefore have a responsibility to ensure all students, regardless of background, develop the skills needed to work confidently and critically with AI.

Highlighting real examples from The Open University, including AIDA and AIMWA, he showed how AI‑enabled support and learning‑material production are already reshaping workflows. These innovations demonstrate a wider shift: from expecting staff to do all the work, to preparing them to run, govern and improve the AI systems that now underpin modern knowledge work.

Ultimately, Prof Domingue offered a clear message: by 2036, thriving in the workplace will depend not on one‑off performance, but on the ability to build repeatable value with AI. The universities that prepare students to be drivers, rather than passengers, will shape not only employability but also the fairness and inclusivity of our future society.