When Will They Learn? Episode #3 When Will They Learn?
12 snips
Mar 5, 2026 They debate the use of AI for homework and marking, weighing workload savings against risks like standardisation and errors. They explore practical ways teachers currently use AI and how to constrain tasks to limit misuse. They unpack what differentiation really means, its evidence gaps, and unintended consequences for expectations. They also discuss offers of initial teacher education places to applicants with low ATARs and what that might imply for teacher readiness.
AI Snips
Chapters
Transcript
Episode notes
Avoid Relying On AI To Mark Homework
- Avoid using AI to auto-mark homework; prefer in-class assessment you can observe.
- Caiti and Greg note homework marking is a Sisyphean task and AI may mark AI-generated work, reducing validity.
AI Missed A Physics Calculation Error
- Greg used AI to check handwritten physics solutions and it initially marked a mistaken step as correct.
- He later found he'd multiplied instead of divided by mass, showing AI can miss calculation errors and needs teacher checking.
AI Confused Circle And Parabola In A Math Question
- Caiti asked AI to write a quadratic question but it produced a circular hamster wheel modeled with a parabola, mixing up circle and parabola.
- She shows AI can misinterpret mathematical structure despite plausible wording.