Anyone who has never been tempted to use ChatGPT to cheat should cast the first stone. Since the launch of this generative AI at the end of 2022, many students have used this language model as part of their homework. However, a teacher has developed a method to spot fraud with formidable efficiency.
A very effective approach
Lucas Markarian, mathematics teacher in a middle and high school in Marseille, also works on TikTok under the pseudonym @lucasmaths4. His videos are so successful that his account now has more than 700,000 subscribers.
In a video spotted by Le Figarohe reveals his almost infallible technique:
It’s simple, when the students give me a homework assignment, I tell them ‘take out a sheet, homework on the table now’ and I ask them the same questions as in the homework assignment, explains the mathematics teacher. This technique is very effective. If they all get 19 on the homework assignment and 4 on the table assignment, then they cheated.
If he is not opposed to AI as such, he deplores the fact that students do not understand the exercises, because that remains the final goal. The professor adds: “Those who use ChatGPT often wax lyrical, adopting math teacher formulations. They think they are Baudelaire, he quips. That doesn’t make them bad people, it just means they’re not yet smart enough to know how to cheat.”.
ChatGPT spoils the atmosphere in class
Regardless, the arrival of ChatGPT creates new challenges and tension in the world of education. We recently spoke to you about his parents who decided to continue their son’s establishment. The latter decided to sanction him because he used generative AI for an assignment. They believe that this technology is not prohibited by the regulations and that the punishment risks costing him studies at certain prestigious universities. More information here.
Finally, note that, according to a recent survey carried out in the United States, 90% of teachers say they are more suspicious than in the past of their students, whom they often suspect of cheating with AI. This is therefore clearly not likely to create a climate of trust and respect between adults and children. We invite you to read this very informative study here.
