ON A Monday afternoon in May, a final-year student, fresh off the Texas A&M University-Commerce graduation stage, received a shocking email. 鈥淭he final grade for the course is due today at 5 p.m.,鈥� it read. 鈥淚 will be giving everyone in this course an鈥� incomplete.鈥�
According to a report in the , agricultural sciences professor Jared Mumm had run his students鈥� essays through the AI tool ChatGPT, which had detected its own use in the work 鈥� an offence that warranted a zero on the assignment. But in reality, it was ChatGPT that was the plagiariser, incorrectly claiming credit for several student-written papers.
Advertisement
Mumm鈥檚 concerns had some merit, though. to write essays or . Some teachers are worried that academia may be crumbling under the pressure of 170 trillion text data parameters packed into a tempting AI chatbot.
The education sector is trying to get a handle on this technological maelstrom. Educational institutions have . Governments are building . Plagiarism detectors GPTZero and TurnItIn attempt to detect AI-facilitated writing, . Meanwhile, AI companies are drowning teachers and government regulators in new and more accessible chatbots.
However, I believe the main problem with the response to such generative AI in education is that the focus is on regulation, not reformation. We should instead be asking: why is it so easy to use these systems to cheat? And why aren鈥檛 we using them to learn?
Traditional learning is extremely linear: memorise content, receive assignments, take a test, rinse and repeat. Students are constantly absorbing rote methods of executing a task or calculation, as opposed to discovering answers by conceptualising the work. As a result, the right answer takes priority over the right thinking, and AI chatbots can get to that right answer very conveniently.
Recognising this, in 2021, I founded , a civil society group mobilising citizen, private and public sector stakeholders for responsible technology. As a youth-led organisation that now has more than 1500 members, one of our main focuses is technology in education. We built an application that uses GPT-4 to generate courses on any topic a student or teacher chooses. The like a more streamlined version of a standard Google search, ensuring that students are presented with factual information and reducing the effect of AIs 鈥溾�� and producing incorrect information.
To supplement the app, our team also designed an assessment system for teachers centred around class discussions and students鈥� verbal explanations of content, using ChatGPT and our app as an educational aide. We now have just over 2000 young people and teachers across the US, sub-Saharan Africa and South Asia plugged into our platform. Girls in Ethiopia have used it to understand menstrual health, while neurodiverse teens from low-income New York households used it as an English comprehension tutor. After four months, we had some students jumping one or two grade levels in core subjects when benchmarked via national exams. Perhaps most impressively, they were doing this with limited access to textbooks or teachers with specialised degrees.
So the discussion about how to integrate AI and education isn鈥檛 just for big tech and policy-makers, it is one for us all. Because if we build and use AI effectively, we can create an education system where students are assessed on the quality and depth of their knowledge, not the content of an exam. This won鈥檛 be the pathway to an easy A, but to an accessible and more effective education.
Okezue Bell is an AI activist, research student and founder of Fidutam, a civil society group
