Students who don’t use AI are ahead on every measure
Every year, the OECD surveys students around the world for the Programme for International Student Assessment. The 2025 results were just released today. [OECD; OECD, PDF] Around the world, results have gone down over the past ten years. But PISA includes an AI measurement. How do students do when…
Every year, the OECD surveys students around the world for the Programme for International Student Assessment. The 2025 results were just released today. [OECD; OECD, PDF] Around the world, results have gone down over the past ten years. But PISA includes an AI measurement. How do students do when they use the chatbot in school? It turns out that winners don’t do chatbots. The graphs show that students with little or no AI use trounce the students who do use the AI: In these scenarios, students who do not use AI outperform those who do, on average, by around 20 points — equivalent to being around one year of schooling ahead of their peers. The summary subheading claims “the relationship between AI use and student performance is complex”. What that means is the AI avoiders do much better. The kids who do use AI do at best about the same. Some do slightly better — but also just happen to be richer. The report’s intro notes the obvious: In the same way that we do not become fit by watching sports but by doing sports, learning does not occur through the consumption of content, but as a productive cognitive struggle of the mind with new material. Of course, the report then tries to both-sides: If we get this right, AI becomes a scaffold, not a crutch. A tool for thinking, not a substitute for it. But I suggest there’s one responsible approach to the homework machine in education: don’t. Winners stay left behind. Video — PodcastSource: Pivot to AI — Published — Category: Business