Education is on the frontline of the AI revolution
You’ve probably heard a lot in recent months about students using ChatGPT to write their school essays, and the scramble to upgrade plagiarism checkers to spot AI-generated content.
It makes for dramatic headlines, but in a few short months, many in the education sector have evolved their approach, seeing AI as a genuine tool to improve the process of learning.
Recently, I was back in study mode myself as I undertook an AI For Business Specialisation via the Coursera platform from the University of Pennsylvania. That exposed me to the thinking of Professor Ethan Mollick, the renowned Wharton School resident expert on innovation and entrepreneurship at the university.
Chandima Kulathilake is an Account Technology Strategist for Microsoft based in Wellington.
Instead of trying to keep AI out of the classroom, Mollick is instead requiring students to embrace it to develop their papers and projects.
“It completely changed how the class worked,” Mollick told a digital learning summit earlier in the year.
Projects that would only have been presented in theory form as paper-based outlines, progressed rapidly to fully-fledged prototypes as a result of students using AI.
“I think we are going to be fine” Mollick says of AI.
I think there are tools that will make education better and more effective. There’ll still be a need for classrooms, at least in the medium term. We’ll figure this stuff out.
I tend to agree with him. In my role at Microsoft looking after education sector clients, I see huge potential for the role artificial intelligence can play in improving the process of learning.
As Mollick puts it, based on his experience teaching with AI in the classroom, we can anticipate three things happening:
- Expect more from students: Given licence to use AI, students should produce better essays, complete with the prompts they used to extract answers from ChatGPT, and put together projects faster, and to a higher standard.
- Better pedagogy, less work: The science of education is far from perfect and we do things in education that haven’t really changed in hundreds of years, and which probably weren’t that effective to begin with. AI gives us the tools to personalise education, tailoring it to the specific needs of a student. It also allows educational institutions to assess the efficacy of the teaching approach more accurately they are taking.
- A completely new world of ed tech: We haven’t scratched the surface of how AI will change the face of education, in terms of formats of lessons and classes, improvement in self-tutorial programmes, and enhancing life-long learning
As I talk to people in universities and at other education institutes around the country, I’m detecting a lot of interest in AI along with a nervousness about what it means for the way students are taught and educational programmes are run. This is only natural; the technology is moving so fast.
While it's easy to focus on the AI tools that are emerging and which education sector CIOs around the country are being asked to assess, vice chancellors, heads of department and senior staff really need to be considering their approach to AI and adopting a framework to deploying it, even if it starts with small-scale pilot projects.
That framework needs to incorporate a responsible approach to deploying AI, guided by well-developed guidelines and following best practice when it comes to the implementation. That will naturally encompass things like protecting privacy and data security, minimising bias, and ensuring human oversight of AI-powered decision-making systems. As a member of AI Forum working groups, we’ve been developing a set of toolkits that any organisation can use to add some tried and tested methodology to this process.
The power of personalisation
It’s important to carefully consider the use cases for AI in education, its challenges and limitations and its capacity to disadvantage a group of people. A good AI risk assessment will put that to the test.
The education sector can draw on sets of generalised open data already in the public domain, as well as individual, identifiable data from students. Both are potentially very valuable to improving educational processes and learning outcomes. But education practitioners need to be very clear about what they are using those data points for.
AI capabilities, when implemented with good oversight, I see as an opportunity to actually help those who, for various reasons, are already disadvantaged.
If we can get it right, I think that personalisation AI could bring to education has the potential to be incredibly positive. The simple fact is that a single lecturer or teacher just doesn’t have the capacity to attend to every student’s needs, and to pay particular attention to the ones that need a bit more help.
There's a massive opportunity for us to think quite differently about education as we look to the future. I’ll give the last word to Professor Mollick:
It’s not going to be impossible to adjust, but we don’t know what that future looks like, and the best thing you can do right now is stay nimble by trying these things out yourself.
Science and Technology Journalist