Machine learning: universities ready students for AI revolution

Sydney’s move towards programmes focused on problem-solving designed to address rise of artificial intelligence and automation

August 3, 2017
Giant robot legs
Source: Getty
Robot masters: graduates need the problem-solving skills to cope with the rise of artificial intelligence and automation

A vice-chancellor’s call for universities to train undergraduates “to tell the machines what to do” has rekindled debate about how higher education institutions can best prepare their students for the jobs of the future.

Michael Spence outlined plans for the University of Sydney to move towards offering four-year degrees with a greater focus on problem-solving and cultural competency as sector leaders around the world debate whether the rise of artificial intelligence and automation will require providers to prioritise specialist skills in areas such as coding, or broad knowledge that will allow graduates to adapt to a changing workplace.

The shift towards longer degrees also runs counter to the push in the UK for more two-year degrees, designed to allow students to start their career more quickly and more cheaply.

In an interview with Times Higher Education, Dr Spence outlined how Sydney had streamlined its 122 degree programmes – a portfolio based on the supposition that “if you enter a narrow tube that has a job name at one end, at the other end you’ll plop out into the job” – to just 25.

The rise of AI means that such jobs “may not exist by the time you end up there, or at least won’t necessarily have any longevity”, Dr Spence said.

Next year, Sydney will introduce a four-year bachelor of advanced studies degree, combining specialism in a discipline with the option of taking a second “major”, foreign language training, international mobility, and cultural competency training.

The signature element of the programme is an extended problem-solving project, where multidisciplinary teams of students work together to tackle a real-life challenge set by a business or charity.

“The skills that machines won’t have particularly well – creativity, interpersonal skills, the ability to think laterally – [these are] the kinds of skills that you need to tell the machines what to do, [and] university education is going to need to be shifted more in that direction,” Dr Spence said.

Dr Spence, formerly the head of the social sciences division at the University of Oxford, admitted that Sydney had been “very nervous” about the change, since degrees would map less closely onto potential jobs. There was an expectation that many parents, particularly in overseas markets, would prefer students to take a degree in what Professor Spence described as “table and chair science” with a view to “going into the table and chair industry”.

But the response has been positive, he said, adding: “People get that you need a different kind of education for a much more uncertain world of work.”

This is just one of the approaches being taken by universities to prepare students for jobs that don’t exist yet. Other institutions, such as the universities of Edinburgh and Adelaide, and University College London, have focused on developing similar sorts of skills to those that Sydney is cultivating by getting students involved in conducting research.

Leaders of Asian universities in particular, and some US institutions, have suggested that institutions may need to prioritise software and data skills in response to the so-called “fourth industrial revolution”.

But Asia and parts of Europe have also seen a revival of liberal arts degrees, focused on equipping students with broad knowledge across a range of disciplines.

Mike Sharples, chair in educational technology at the Open University, said that developing graduate attributes such as critical thinking, creativity, global awareness and networking skills were key tasks for higher education institutions.

He argued that graduates would need to know how to work with AI systems, meaning that they would need some “quite specific skills” to help them understand how AI systems operate.

“This [the combination of these approaches] is something that universities really have got to get to grips with,” Professor Sharples said. “It’s done in pieces in some universities, whether through career planning or through introductory courses, but I don’t think that any university does it in a really systematic way.”

Sector leaders predict that universities may also need to develop much more meaningful lifelong relationships with their students: providing an initial higher education experience, and retraining at regular intervals during a graduate’s career as the jobs market evolves.

Simon Marginson, professor of international higher education at UCL, said that he “worries to see the same old debate between liberal immersion in knowledge – as generic training in self-developing thinking and creativity – versus curricula attuned to industry specifics”.

“Obviously both are needed, if we want self-determining people who understand the context and the processes, and can both lead and work to a set agenda,” he said.

Growth predicted for ‘big-picture hybrid’ degrees in ‘smaller’ sector

Universities are likely to get smaller and focus more on “big-picture hybrid” degrees in response to the impact of artificial intelligence and automation, a researcher has claimed.

Peter Murphy, adjunct professor in the School of Humanities and Social Sciences at La Trobe University, and the author of Auto-Industrialism: DIY Capitalism and the Rise of the Auto-Industrial Society, told Times Higher Education that “the era of the mass university is over” and that universities “will have to return to their traditional mission of educating exceptional talent”.

Higher education expansion had been driven since the 1970s by the demand for white collar workers across the Western world, according to Professor Murphy, who said that many of these jobs were “now disappearing”.

“Universities won't disappear, but they will shrink in size,” he predicted. “Some will go out of business entirely; vice-chancellors are in self-denial about this. They are resisting the number one imperative of the age: get smaller.

“Many more young people in the future will go from secondary school straight into the workforce. For one-third to half of students today, the return on their investment in a university education is poor.

“Tertiary-level self-learning is increasing; competency-based qualifications are growing, they already dominate in IT. In place of a taught course, you simply sit an exam to determine what you know.”

In this environment, some “old demanding disciplines”, such as engineering and political science, will continue to do well, Professor Murphy said, because “their graduates can’t be replaced by an algorithm”.

But the “big losers” would be “studies-style” courses such as media, business and environmental studies, Professor Murphy said, with jobs linked to these courses gradually being automated.

“On the upside, growth will occur in big-picture hybrid courses,” Professor Murphy continued. “These meld technology, mathematics and social science. Amazon is the paradigm of the auto-industrial employer: it wants high-level graduates who can work across economics, technology and data science.

“This is not a mass job market. So universities will have to return to their traditional mission of educating exceptional talent.”

Professor Murphy added that while the workplace was becoming “more entrepreneurial”, universities “think only in terms of salaried work for graduates” and were “very uncomfortable with graduates starting up small businesses”.


Print headline: Machine learning: readying students for the AI revolution

Register to continue

Why register?

  • Registration is free and only takes a moment
  • Once registered, you can read 3 articles a month
  • Sign up for our newsletter
Please Login or Register to read this article.

Related articles