Skip to content
All tags

#ai-and-learning

3 posts
learningdeep-diveWriting Articles That Teach

What Research Measured on AI-Written Teaching Materials and AI Tutors: Content Error Rates, Human Screening, Student Ratings, and Tutoring Experiments

When GPT-3.5 wrote solution hints for algebra and statistics problems directly, both the answer and the steps were wrong for 32% of the problems; after a human screened every problem, no difference in learning was detected between those hints and hints written by human tutors. Randomized experiments on AI tutors point both ways: an unguarded chatbot made high school students do worse on an exam, while the positive results mostly came bundled with worked solutions written for each problem, a teacher in the room, or extra class time. None of the studies tested the same content again after a delay.

#learning-science#ai-and-learning#ai-tutor#self-learning#course-guide
learningguideEnglish Speaking Practice

How to Practice Workplace English Speaking: Shadowing, Scenario Drills, AI Apps, and Meeting Phrases

30 minutes a day for 12 weeks — combine shadowing, scenario practice, and AI apps to go from 'I understand but can't speak' to holding your own in meetings.

#english#speaking#self-learning#ai-and-learning#language-learning
learningdeep-diveLearning How to Learn

AI Makes Things Smooth Exactly Where They Should Be Hard: What Generative AI Does to Learning

The single most-cited meta-analysis on ChatGPT in education (g = 0.867, ~500k views) was retracted by Nature in April 2026. But the positive finding was not overturned — the issue is that it measures performance while the AI is available. Bastani's PNAS RCT measured something else: +48% accuracy during practice with GPT-4, then 17% below never-users once access was removed.

#ai-and-learning#learning-science#llm#metacognition#self-learning