Mr Barton Maths Podcast #217 AI in Education with Adam Boxer
12 snips
Mar 25, 2026 Adam Boxer, a science teacher and education writer who co-runs Carousel Learning, brings a sharp, critical eye to AI in schools. He discusses AI marking, hallucinations and accuracy limits. He weighs workload savings against lost intellectual labour. He examines student-facing AI risks, ethical scraping concerns, and realistic, constrained uses that keep teachers in control.
AI Snips
Chapters
Transcript
Episode notes
Don't Auto Mark Student Work Without Safeguards
- Avoid putting student-facing AI that auto-marks directly in front of pupils unless accuracy and pedagogical safeguards are proven.
- Prioritize self-assessment routines and only automate where you can guarantee high-fidelity marking and maintain learning behaviours.
AI Changes Incentives And Risks Skill Atrophy
- AI tools change incentives and remove cognitive labour from students and teachers, which can harm long-term skill development.
- Boxer argues resisting student-facing AI preserves students' need to learn and self-assess rather than take the easy route.
Narrow Homework Beats Open-Ended Study
- Boxer contrasts two homework approaches: open resource learning versus targeted flashcards tied to a follow-up assessment, saying narrower, specific tasks work better.
- He argues specific, routinized instructions reduce students choosing ineffective study strategies amplified by AI.