Small Steps towards a Digital Future: A Conversation with the University’s New AI Chair

Looking around the world now, it feels like every area of life is slowly being overrun by the growth of AI. From simple searches to new computer models, the landscape of our education is changing week to week. Grappling with these changes, the University of St Andrews has created the Johann and Gaynor Rupert Chair in Artificial Intelligence. Sponsored by the Alfred Dunhill Links Championship, the appointment placed Lars Kotthoff at the helm of the University’s changing relationship with AI. I spoke with Kotthoff to explore how the school plans to move forward in an ever-adapting world. 

Kotthoff acknowledged the difficulties that AI poses to anyone trying to make a plan for the future: “[It] is really challenging, because [AI] changes under our feet as we move across it.” Still, he feels that there is something the school can do, must do, to keep pace. 

What will these plans look like? Kotthoff, whose official job as Chair begins when his year-long sabbatical ends this summer, confessed that the rapidly changing nature of the AI world makes it hard for the school to make promises. “If you put something together at the beginning of the semester, that doesn't mean that it's still valid [in] two weeks,” he said. 

It does no good to ignore AI completely. The job market might require skills relating to AI, and it’s important that graduates can determine if it works, and more importantly, when it fails. “[After graduating] you will work in companies and for institutions where you might be expected to use these systems,” Kothoff explained. “Or at the very least, there will not be as much pushback.” In Kotthoff’s view, to ignore AI would be to set students up for failure in the long term: “At the end of the day, it’s about getting the job done right.”

To do so, Kothoff explained the inadequacy of university-wide policies. For him, it’s important to give students the skills to engage critically with AI, while allowing each department and each professor to establish their own rules. The goal isn’t to “forbid that students use [AI], but essentially to show students that this is where it can go wrong,” he said. 

“I'm a fan of leaving it up to the particular professor or the instructor because it does depend on the context,” Kothoff mused. “In some sense, in some classes, it might make sense to just completely forbid it, because [...] you know, we're philosophers, and we want to write things down on paper and really think about things, whereas in other contexts [...] we're computer scientists and you can generate this code.” Kothoff plans to adapt course content with these considerations.

Kotthoff’s role is not to upend the school’s entire education system – he wants to help find ways to engage with AI that don’t ignore its environmental impacts, but also don’t ignore the fact that it’s is becoming increasingly ingrained in our lives. 

Kothoff also sees scope for AI to help the University itself. “The hope is that AI could eventually not only enable better teaching in the classroom, but also enable [the University] to get better research grants and to make it easier to write proposals and compete for funding,” he explained. “I'm really thinking of this very holistically, as in, how can this benefit the University, and then, of course, also the community beyond the university as a whole.’

There is no easy answer to the question of AI in schools, especially with the number of factors at play. Rules are important to prevent the abuse of AI’s power, but Kotthoff believes that there should also be guardrails in place. “Obviously, we don't want the students who can afford the super duper plus subscription to ChatGPT, to have an advantage over the students who can't.”

So Kotthoff has hope for the University and its ability to engage with the future. "We do need to be careful with these kinds of knee-jerk reactions right where, oh no, we're allowing it, no wait, we're forbidding it, and then applying this to everybody, regardless of discipline,” he said. “I would hope that St Andrews is better than that.” 

Cautious optimism, but also an open mind — this, Kotthoff believes, is how we should see the University’s attempts to get a head start on AI.