AI: transformative for university teaching or ‘CV bullet point’?

Transformation brought by generative AI will mean better research papers and researchers being judged on science rather than how they write, while bringing a focus in universities on teaching students how to prompt AI, according to a computer scientist who leads the University of Potsdam.

Oliver Günther, the president of Potsdam and former professor of information systems at Humboldt University Berlin, recently warned Times Higher Education’s World Academic Summit, hosted by the University of Sydney, that “many of our colleagues are underestimating how AI is going to change teaching and research and administration”.

“For research, many people haven’t understood it will change the way we write papers, change the way we distribute our time between doing the actual research and writing it up,” Professor Günther told THE.

“It will lead to better papers because many scientists have trouble using the kind of prose that could reach a wide readership.”

Generative AI, he predicted, will also boost researchers who publish in English but for whom English is not their native language.

“I think the advantage of having English as a native language is going to go away – which is good for everybody,” said Professor Günther. “So science is not being judged on how it is written up; that will change.”

Another speaker at the summit, edX founder Anant Agarwal, told the event that universities were failing to keep pace with the corporate world in appreciating that “prompt engineering” – setting guidelines in interactions with AI and “asking questions in a way you get answers you’re looking for” – will, in future, be “the single biggest skill everybody has to learn”.

Campus resource: Can we spot AI-written content?

Professor Günther said: “You are basically an educated expert proofreader of the text that comes out of the machine – that’s what we have to teach.”

“Any faculty that has any kind of responsibility for their students will understand that to teach computer scientists [for example] how to write computer science papers has to be different now than it was two years ago” before the advent of generative AI, he continued.

Key questions were now, Professor Günther said, “how do you prompt the generative AI, not only for the text but also for the diagrams…that is what we have to teach, then we have to teach how to judge the first draft of what comes out of the machine”.

But the focus should always remain on “what we want to achieve as a university, what we want to teach our students, rather than starting from the technology,” he stressed.

‘It is just a professional skill, not a core part of our curriculum’

Graduates’ ability to use generative AI will be a CV bullet point on a par with “can use Microsoft Excel” rather than transforming what universities teach, according to a specialist in using AI and digital technology in the classroom.

David Kellermann, senior lecturer in UNSW Sydney’s School of Mechanical and Manufacturing Engineering, built a conversational agent, or bot, to assist students as early as 2017 and attracted the attention of Microsoft for his linked use of Teams as a learning platform.

One “knee-jerk reaction” to the advent of generative AI, he told Times Higher Education at the recent World Academic Summit, was to argue that universities must now teach students to be “masters of the AI – we have to learn how to drive the AI; because the AI knows everything and can do everything the new disciplinary frontier is prompt engineering."

That was, he added, an “excessively optimistic but simultaneously dystopian view. What it does is it forgets the place of human thought: it thinks that a large language model that’s basically trained like a pastiche of the internet is able to replicate it [human thought].”

“What generative AI actually does is it threatens the mechanism we have used to assess students’ learning,” meaning “dramatic impact on remote digital assessment” and a shift to in-person assessment, he said.

“The real reaction [to AI] we’re going to see is that universities are going to realise or remember what their value proposition is – it is to offer a formative experience,” predicted Dr Kellermann.

He continued: “Your competency in using generative AI, or prompt engineering as people are calling it, the place of that particular skill on your résumé will be a bullet point in professional skills, and it will sit there right after ‘can use Microsoft Excel’.”

As digital computer-aided design replaced the pencil and drafting board, there are “certain things generative AI will replace,” said Dr Kellermann. “But it’s not going to be a core part of our curriculum. It is just a professional skill. Yes, it’s going to be embedded in everything, but I think we are getting carried away in the idea of how it’s going to change that arc of learning that is part of the tradition of universities.”

john.morgan@timeshighereducation.com