ChatGPT and other forms of generative artificial intelligence (AI) remain a concern for those in the education system, especially when it comes to understanding how its capabilities can be misused by students and the long-term impact this would have on their education and careers.

It is almost impossible to open any discussion around education and AI without ChatGPT being mentioned. The tool has caused controversy and debate since it rose to immediate fame towards the end of 2022, and today many educators and institutions remain uncertain as to the future of studies with AI, or without it.

For Anna Collard, senior vice-president: content strategy and evangelist at KnowBe4 Africa, it is a layered and challenging situation that can be both of huge value to the organisation and an immense risk.

“The problem with these language models is that they feed on existing data to create the answers, to build the reports, but this introduces the risk of inaccuracy,” says Collard. “Then, if the data is not checked or managed, it creates more inaccurate data that is being fed back into the model to create even more inaccurate content. A vicious cycle of fake news and false information that will perpetuate errors and impact the value of education. Although this may only be a short-term issue, as within the AI space, models have been developed to improve on accuracy over time.”

It is a risk recently unpacked in a Wired podcast on how educators are now seeking “unhackable” assignments that are almost impossible for AI to write. Yet, this discourse raised another consideration, one that has not been missed in the uproar – is teaching falling further and further behind the real world?

“Perhaps it is time for the concept of generative AI to be looked at in the same way as the calculator,” says Collard. “In the past, the calculator was considered this hugely concerning technology that would completely destroy education, but now it is part and parcel of the education system and on every child’s stationery list.”

It is a view shared by a recent analysis in the New York Times which suggests that the tool should become part of the school curriculum and teaching methodologies. After all, banning ChatGPT will not work; people will use it anyway, and refusing to accept that it is a useful tool for students is the education equivalent of sticking your head in the sand.

Instead, it should be treated just like the calculator: as a tool that students can use to manage their learning and that educators can use to enhance their teaching.

“When the calculator became popular in the 1970s, people were asking if it was a good thing or a bad thing; if it would prevent people from making the calculations in their heads,” says Collard. “And yes, we cannot quite remember as much at one time as we used to, but the calculator is just a tool that requires users to embrace critical thinking. This is the same view that can be applied to ChatGPT: that we teach children to fact-check the AI and to remember they cannot trust everything it says. This will, again, put the onus on critical thinking.”

This shifts the dialogue around ChatGPT away from a knee-jerk reaction of dismay and fear – a reaction that is valid to a certain extent – towards one that is looking at how the vagaries of AI can be harnessed to fundamentally improve how children and students learn.

“If students are constantly challenged to check the data, ensure the information is right and find the truth within the AI, then they are learning; they are applying critical thinking to the topics and discussions,” concludes Collard. “It is a powerful way of embedding the learning within their minds and challenging them to stop taking anything at face value. And this can have far-reaching benefits beyond just the classroom.”