AI can produce and grade AP‑level lessons and quizzes, provide individualized remediation, and enable small or niche high schools to offer advanced and vocational courses without large specialist staffs. Teachers would shift from primary content deliverers to inspirers, moral guides, and supervisors of agency.
— If AI can reliably teach advanced high‑school subjects, it changes access to college‑level preparation, alters staffing needs, and raises questions about assessment, oversight, and the civic role of secondary education.
Arnold Kling
2026.05.14
80% relevant
The article documents hands‑on development and prompting of a Claude‑based tutor and discusses pedagogical choices (application vs. 'explain in your own words') and an architectural behavior (default to prolong conversation). That maps directly to the broader idea that AI tutors are entering formal education (e.g., AP courses) and raises deployment issues about student experience, pacing, and question design.
Arnold Kling
2026.05.06
75% relevant
Arnold Kling’s Claude 'skill' is an example of a modular AI tutor applied to a civic subject — like proposals for AI‑tutored coursework — showing how lightweight, portable lesson packages (URL: arnoldkling.com/apps/pst.skill) can deliver structured, Socratic political‑psych instruction outside schools and publishers.
Arnold Kling
2026.05.03
60% relevant
Mark McNeilly's proposal that students should document prompts, outputs, and their reasoning ties to existing discussion about AI tutoring and classroom integration (how to make AI a learning partner rather than an answer machine), a specific education‑policy thread exemplified by 'AI‑Tutored AP Courses'.
Arnold Kling
2026.04.10
100% relevant
Author claim: "the AP curriculum really would be easy for an AI to help teach" and examples of AI polishing lectures, making practice problems, and evaluating student work.