Overview of Rewind: How AI is fueling an existential crisis in education
This Verge Decoder episode looks at how generative AI is reshaping education far beyond the familiar concern of students cheating with ChatGPT. Host Nilay Patel speaks with McGill educational technology researcher Dr. Adam Dubé and multiple teachers to explore the deeper issue: AI is challenging what schools are actually for, how learning should be measured, and whether current education systems are rewarding real understanding or just polished outputs. The episode argues that the biggest problem may not be AI itself, but the way schools are being pressured to adopt it without clear evidence, coherent policy, or attention to student learning.
Main Themes
AI is creating an existential question for education
- Teachers are not only worried about plagiarism and cheating.
- Many are asking a more fundamental question: what is the purpose of school if AI can produce the work?
- The episode frames generative AI as a philosophy-of-education problem, not just a classroom-tech problem.
“Digital natives” is a misleading idea
- Dr. Adam Dubé explains that kids growing up with technology are not automatically skilled at using it for learning.
- Familiarity with phones, YouTube, or games does not mean students understand:
- how to evaluate AI output
- how to reason with tools critically
- how underlying systems work
- The episode argues that schools often overestimate students’ tech literacy.
AI changes how students interact with knowledge
- AI tools are increasingly being used through natural language, which makes them feel intuitive.
- But the podcast stresses that being able to ask AI questions is not the same as understanding its answers.
- Students may trust AI-generated explanations without having the background to judge whether they are correct.
What Teachers Are Experiencing
Mixed reactions, but a lot of frustration
The episode features several educators with different views:
- Some teachers see AI as helpful for saving time on lesson planning, email, or drafting materials.
- Many others say it creates more work, not less, because AI output often needs heavy correction.
- Some educators see AI as actively harmful, especially when it replaces reading, thinking, writing, or subject-specific rigor.
Teachers feel students are under pressure
- Several teachers note that students are not simply being lazy or malicious.
- They are often responding rationally to:
- heavy workloads
- grade pressure
- financial aid concerns
- job/internship anxiety
- AI becomes appealing because it helps students finish tasks quickly, even if it undermines learning.
Autonomy matters for teachers too
- Forcing teachers to use AI can be demotivating.
- The episode emphasizes that educators dislike being told to adopt tools that reduce their professional judgment or control.
Research and Evidence Highlighted
Cheating is real, but not the whole story
- Dubé cites research suggesting about 10% of students report using AI to generate entire assignments.
- More common uses include:
- explaining concepts
- brainstorming ideas
- summarizing text
- editing writing
- The episode suggests that while outright cheating matters, the broader issue is how AI changes learning habits.
AI can reduce memory and retention
- The episode compares AI to calculators:
- tools can help with output
- but if they do the thinking, students practice less
- Dubé points to research showing that people who rely on AI-generated work may:
- remember less
- reflect less deeply
- build weaker long-term knowledge
- A cited MIT study found students had poor memory for essays written with ChatGPT assistance.
Hallucinations are especially dangerous in education
- A historian interviewed in the episode describes machine translation software that inserted fake sentences and paragraphs into historical documents.
- The result: AI created more work and cost more than hiring a human expert would have.
- In fields where accuracy matters, AI’s tendency to “sound right” can be especially misleading.
The Big Structural Problem
Schools reward products, not processes
- A major argument in the episode is that education often evaluates the final assignment, not the learning process behind it.
- That makes AI especially tempting because it can produce a convincing final product.
- But if the student didn’t actually do the thinking, the grade may not reflect real learning.
Current incentives may push students toward AI
- Students are rewarded for completion, speed, and performance.
- Teachers want genuine skill-building.
- AI fits the grading system better than it fits the educational mission.
There is no consistent policy response
- School systems are reacting in fragmented ways:
- outright bans
- “AI everywhere” adoption
- cautious experimentation
- Policies vary by district, community attitudes, leadership philosophy, and budget pressures.
- Administrators may even support AI for cost-saving while discouraging students from using it.
Notable Insight
“What are we even doing here?”
This repeated question captures the episode’s central anxiety: if AI can draft, summarize, translate, and even tutor, schools must decide whether education is about producing acceptable work or developing human expertise.
Key Takeaways
- Generative AI is not just a cheating tool; it is forcing schools to rethink the meaning of learning.
- Students are not automatically better at using AI than adults just because they grew up with technology.
- AI can help with some workflows, but it often introduces errors, weakens retention, and can undermine expertise.
- Teachers are split, but many feel AI has been imposed on them without enough evidence or autonomy.
- The core tension is between speed/productivity and deep learning/process.
- The episode’s final message: education may need fewer pressures and better-designed incentives, not just more AI tools.
Bottom Line
The episode argues that AI in education is less about whether students will cheat and more about whether schools still know how to value real learning. Generative AI may be efficient, but the show suggests that efficiency alone is a poor substitute for curiosity, practice, judgment, and understanding.