in the classroom
company , the most important thing that we can do is to recognise and understand the potential for bias and ensure that we ’ re keeping it in mind , and designing and implementing in a way to minimise that bias .
We know that AI can be tremendously beneficial in unburdening individuals from some time-consuming and often menial tasks . Can you highlight some of the ways it can benefit the sector and people working within it ? I think a good example of this could be around marking . I think everyone who ’ s involved and every teacher I ’ ve spoken to really hates marking . It ’ s time-consuming , it ’ s monotonous , and it ’ s also quite hard , I think , to give meaningful feedback . But also , of course , it ’ s important . You can ’ t really avoid it . It ’ s a key part of being in education . When we think about marking and the way AI can help is it can take away some of the most menial aspects of marking . If you have short answer questions and you have students answering the same question time and time again , you can use AI to speed that process up , but you actually still answer the question and actually have more time to give feedback around that .
I would say when it comes to things like academic integrity , AI can also be used . A big part of academic integrity is about building a body of evidence , of course , when you ’ re investigating things and , again , that can be incredibly timeconsuming . But using AI can really speed up that process of collecting separate pieces of student samples across a course or a cohort and take a job that would be potentially taking , ultimately , lots of hours down into one of minutes .
It ’ s about reducing the time you spend on those low-impact activities and spending it instead on more high-impact things like speaking with students or providing feedback .
AI is being deployed to not only identify cheating , but also teach how to avoid it . Can you elaborate on this ? This comes back to , what is the fundamental question of why do students cheat ? Ultimately , there are a host of reasons and no one single reason , and this very much depends on the student , the conditions , et cetera . But there could be a lack of understanding , it could be that they have a lack of support around their learning , there could be a lack of engagement .
By freeing up the teacher from menial tasks , AI can give them more time to spend with the students to provide them with the support that means that they ’ re less likely to engage in cheating behaviours .
Also it can potentially give support to the student through automated feedback or automated means at the time when they ’ re going to be most vulnerable within the learning process . A student is never going to be able to reach out to a teacher exactly when they need them . It could be in the middle of the night , for example , and a teacher can never respond immediately to every request . So AI has the potential , at least , to provide actionable feedback to students in a very immediate way .
The technology is probably not quite so developed there , when you think about AI being able to provide meaningful feedback
You want the AI to support and complement the current human processes .
on unstructured texts , but I don ’ t think we ’ re too far away from students being prompted to review citations that they ’ ve added to their essay , for example , because the AI has picked up that there ’ s a potential mismatch between what they ’ ve cited in their bibliography , or , of course , that they haven ’ t cited anything at all . So that would be one example where AI could really lean in on the academic integrity .
How important is it that AI remains under our control or human-centred , rather than the other way around and why ? I really like this term , human-centred AI . I think it means putting the human user at the centre of all stages of AI , so research development and the actual functional application . I think this , of course , then comes back down to questions around concerns that people have in adopting AI because , certainly around job losses or how it ’ s going to affect their daily life , you want the AI to support and complement the current human processes . You don ’ t want it to replace what they like doing . You want just to give them more time for them to be more effective as an educator .
Humans at the centre of the research and development stage , I think , is crucial to reducing biases that we spoke about . If you don ’ t keep that in mind , then I think you ’ re both going to introduce technology that has more flaws in it , but also is , essentially and ultimately , going to be less valuable for the end-user as well .
So , yes , I think that human-centred AI is going to be key and crucial for how AI is , ultimately , going to be most effective in education .
You ’ ve also got the ideas around the analysis or the better analysis that you can do , the insights you can derive from data that AI can deliver as well . So , doing things that you could never normally be able to figure out and going beyond what ’ s previously been possible .
I think there ’ s this aspect of creating efficiencies for what ’ s currently possible and then going beyond what ’ s currently possible . ■ educationreview . com . au | 19