Teachers have spent the past few years wringing their hands about ChatGPT’s ability to help students cheat on their assignments. Generative AI could write a college essay or answer a math problem in just a fraction of time, making it a tempting shortcut. Professors—accurately or inaccurately—accused their students of using ChatGPT to complete their assignments.
The education company Chegg estimated that 40% of undergraduates around the world have used generative AI in their tertiary studies, with half of that group using a tool like ChatGPT at least once a day.
But education experts speaking at the Fortune Brainstorm AI Singapore conference on Wednesday believe that, almost two years since the release of OpenAI’s ChatGPT, that generative AI can now be an asset to teachers, rather than a shortcut for students.
When ChatGPT came out, the “knee-jerk” reaction among educators was fear that students would start plagiarizing their assignments, said Sonita Jeyapathy, co-director of the Center for Pro Bono & Clinical Legal Education at the National University of Singapore. But instead, “we realized that we could leverage [Gen AI] as well.”
Teachers are asking AI developers for assistance in planning lessons, motivating students and professional development, noted Khairul Anwar, founder of Malaysian edtech startup Pandai.
AI developers are also building apps to help students with their learning. Pandai has developed a chatbot to help students with their homework—but not do it for them.
“It’s designed not to give the answers outright, but instead to give you step-by-step solutions. To ask the students themselves…what do you understand now, and what do you think the next step is?” said Anwar said.
But chatbots are just the tip of the iceberg.
“There's a lot more that's happening in AI beyond just large language models,” said Tim Baldwin, provost of the Mohamed bin Zayed University of Artificial Intelligence. He cited the example of an AI that can be trained on how a student is learning, then cater the curriculum to match his or her strengths, expanding access to a personalized tutoring experience to those who traditionally could not afford it.
Panelists agreed that AI-enabled cheating was not a new phenomenon.
Jeyapathy said it’s natural for students to want to get a better grade in an easier way. She suggested that a student’s motivation level and a teaching institution’s values are more influential on the decision to cheat than access to any particular AI tool.
Anwar suggested that teachers and institutions needed to better highlight the value of education, rather than promoting learning merely as a way to get material comforts. If education is described as a path to a job, a big house and an expensive car, then “the message is that this is just a transaction, [and] obviously students will cheat.”
