← Home
Decoder · May 7, 2026 · 42m

Rewind: How AI is fueling an existential crisis in education

Decoder’s Nilay Patel revisits a fall 2025 episode examining how generative AI is reshaping education. The show explores widespread student use of AI tools, shifting attitudes toward cheating, new research on cognitive impacts, and how children learn from conversational agents. It’s a comprehensive look at the tension between AI’s potential as a learning aid and the risks it poses to academic integrity and deep learning.

This summary was generated from show notes and public descriptions, not from a full transcript review. Details may contain inaccuracies.

Highlights

Widespread student use of generative AI
A majority of high school students report using generative AI for schoolwork, and about a quarter of teens have used ChatGPT specifically.
Students’ shifting attitudes toward AI cheating
Many students see using AI to complete assignments as acceptable, not as cheating, and some argue institutions should adapt to this new norm.
Cognitive effects of ChatGPT use
The episode cited MIT Media Lab research exploring how interacting with ChatGPT changes the way people think and process information.
How children learn from conversational AI
Research from McGill University indicates that children can learn from conversational AI, but the learning differs from human-to-human interaction.
AI reliability and accuracy challenges in education
AI-generated content often contains factual errors, which poses a serious challenge when used as an educational resource.
Was this useful?