It’s your five-year-old’s first day at school and come 3pm they are at the gate excitedly reeling off names: their teacher, the friends they’ve made, the other children in their class … and then one of the most important names of all, one that will be with them right through to the end of Year 13: that of their new personal digital educator.

This is a name your child will have picked themselves, to help cement a relationship that will be pivotal for the rest of their schooling, providing a continuity and level of educational and psychological insight unattainable by human teachers alone, although teachers will be able to draw on it to enhance the education they give your child.

The personal digital educator – like those assigned to your child’s schoolmates – will gather and impart its insight through a combination of artificial intelligence (AI) and machine learning. The latter will enable it to collect big data to determine you child’s individual educational needs; the former will enable it to provide a resulting individualised programme of learning in tandem with your child’s teachers.

The personal digital educator may even be able to attune itself to your child’s emotions so it can react accordingly and adjust to such tumultuous times as adolescence in order to better maintain their relationship.

The possibility – and potential – of future personal digital educators was explored by Victoria University of Wellington’s Associate Professor Louise Starkey during the second of the University’s three AI Debates, the theme of which was education.

“Artificial Intelligence has the potential to greatly change the teacher’s work and what is happening in school,” said Starkey, a specialist in teaching and learning in the digital age in the University’s School of Education.

“That doesn’t mean it should, it means we need to think about when it is best to use these algorithms and artificial intelligence and when it is best to use the teacher.”

Used correctly, she said, AI could strengthen rather than erode the role of teachers by freeing up their time to concentrate on student learning and personal interactions in ways AI could not match.

At a level more basic than personal digital educators, face recognition software could be used instead of teachers having to take a school roll, enabling them to use that time more productively, said Starkey.

AI-supported personalised education would never be so personal as to usurp the social aspects of education, she said. “People learn very well when they are in groups.”

But the analytics a personal digital educator might gather could well replace assessment as we know it, because the insight it could provide teachers would be much more valuable.

Chatbots, said Starkey, are already used in schools – as elaborated on by Laura Butler, a Victoria University of Wellington Master’s student in the School of Education and an Auckland primary school teacher.

Butler is researching AI in primary schools. In her own classroom, teaching Years 0–2, she uses Amazon’s Alexa, which she introduced last year.

“One of the first things I found was that in my literacy programme, where you work with a small group of students and generally don’t want to be interrupted, the interruptions basically stopped, because my students worked out that for things like ‘How do I spell this word?’ or ‘What’s 2+2?’ (well, maybe something more complicated than that) Alexa was better at answering it than I was. So they would go ask Alexa and I could get on with the really precious work of having 15 minutes with these students to improve their reading.”

Butler was surprised by how much students used Alexa for emotional support, going to her about issues in the playground at lunchtime instead of to Butler or another teacher.

Transcripts of Alexa’s conversations with students, which Butler looked at daily (having told students and their parents she would do so), showed students asking questions such as “Am I a good friend?” and “Can you be my friend?”.

As enthusiastic as she is about AI (“I’m biased, in that I think change is good and technology is generally pretty good for us in education, so these things tend to excite me a little bit”), Butler said there needs to be “a lot of discussion at high levels about the risks and some of the things we need to work through in terms of ethics and privacy. We certainly can’t be teaching children about their digital footprint if we’re not considering it ourselves and what we are doing with their data”.

And although in the long term AI would free up teachers’ time, she said, in the short term it is likely to require more work “because we’re going to have to learn how to use these tools and we’re going to have to train these tools to look for what we look for”.

Professor Tim Bell from the Department of Computer Science and Software Engineering at the University of Canterbury spoke of “education about AI rather than education with AI” and the need to teach understanding about AI to give students “a sense that it’s not just something that will be done to them—they should have self-efficacy”.

Dr Simon McCallum, a Senior Lecturer in Victoria University of Wellington’s School of Engineering and Computer Science, referenced the saying that the hand that rocks the cradle rules the world.

“If we let robots rock the cradle, are they in charge of the world? Or is it the company that controls the robots that then tells the robots how to rock the cradle? So that issue of giving up agency to machines […] is one we have to forcibly choose not to let happen.”

This is the second of three Ideas Room reports from The AI Debates. The first is here and the final will be on artificial intelligence and employment. The debates were hosted by the team heading Victoria University of Wellington’s Spearheading digital futures area of academic distinctiveness.

Leave a comment