Skip to content
The UK's only independent think tank devoted to higher education.

What does artificial intelligence (AI) mean for education? (And why do presentations on new edtech usually rely on old tech?)

  • 23 October 2019
  • By Nick Hillman

It used to be thought the modern age would herald meals in the form of small pills, supersonic travel and fewer working hours.

It hasn’t turned out like that. Instead, we rather like slow food. We tend to prefer cheap to super-fast travel. More income has often proved more desirable than more leisure.

In higher education, there’s been a similar debate. The invention of the radio, talkies at the cinema, compact cassettes, video tapes, the internet and MOOCs were all once predicted to transform higher-level learning beyond all recognition – as recalled in this extract from Derek Bok’s excellent book Universities in the Marketplace: The Commercialization of Higher Education (2003).

These things have all had their place in making teaching and learning more efficient, of course – many a teacher, including myself in an earlier career, would have found life even harder without occasionally resorting to a video or two (and, sometimes, the learning was improved as a result). I used older technology, like an overhead projector and even slides too, although most of the time I didn’t use any such aids.

It always bemuses me when, at education conferences, the audience is regaled about the imminent transformation of teaching and learning. I try to absorb all that I am being told but a little voice usually pops in my head to ask: if the changes are so transformational, why are they being explained in such a traditional, almost chalk-and-talk, way? A PowerPoint slide or two is generally the limit.

Perhaps artificial intelligence (AI) is different. Perhaps we are, this time, truly on the cusp of something completely different.

I sometimes doubt this, given my Alexa is much better at making fart noises in response to children’s requests than it is at useful-but-simple tasks like turning on the radio or playing a particular song. Yet perhaps Alexa came to market too soon. Maybe, in a few years’ time, similar technology will have truly transformed all our lives.

I admit to further doubt for a more fundamental reason. In a decade’s time, my younger child will still be young enough to have school homework. Artificial intelligence won’t be able to do it for her if she is to benefit from the process of learning. “That’s true,” say the true believers in AI as an educational tool, “but it will make her teachers’ lives much easier because a machine will mark the work.” Instructors will be able, finally, to have a better work:life balance or to spend their working time completing more useful tasks like providing more one-to-one support.

I’ll set aside that marking is itself part of the learning process for becoming an effective teacher and that such claims were previously made about ‘flipped classrooms’, which have turned out to be largely illusory. But I still don’t buy the wilder claims about AI because another part of the learning process is securing validation and approval from other human beings.

Machines have long marked multiple-choice answers but, if a pupil thinks their essays will only ever be evaluated by machines too, why should they bother trying so hard to make them interesting? If we thought only machines were going to read HEPI publications, it is very unlikely we’d put so much effort into them.

I’ve been thinking about this because of some of the things that students told us in HEPI’s recent research with Unite Students. When asked about different ways to learn, lectures come top.

Why? Not because the learning is better in lectures (it isn’t) but because people want to learn alongside others. When asked what should replace lectures if they were to disappear, technological solutions are much less popular than learning from other humans. 

Indeed, there seems to be a tension in education between the clearly understood benefits of team working and the claims now being made about how machine learning will respond to each of our needs as individuals.

Perhaps, as with previous major technological innovations, for the foreseeable future AI will further improve but not wholly transform what educational institutions do to engage their students.

Leave a Reply

Your email address will not be published. Required fields are marked *