Volume 104 • 2026 • Fall 2026

Generative AI turns up again this issue, but the pieces I keep returning to are less about what the machine produces than about who is in the room deciding whether it is any good. Nabb brings Number Talks into undergraduate calculus and invites a chatbot in as one more voice; the moments that stay with me are the ones where his students refuse a fluent, confident answer — “that’s AI just exhibiting sycophantic tendencies.” Nava Guzmán and Sánchez Aguilar formalize that same division of labor into a six-phase model for analyzing students’ mathematical writing, in which the language model counts and the researchers interpret. I’ve added a short commentary alongside Nabb’s article, because his scoreboard — AI: 1, me: 0 — is more generous than it is accurate.

The rest is ordinary good mathematics teaching. Cox and Harper ask students to build animations in dynamic geometry software, where nothing works on the first attempt and revision stops feeling like a penalty. Taite and DiNapoli hand an Algebra 1 class a yogurt shop and ask for a shelving layout a real business could live with — accessible, appropriately stocked, and answerable to constraints the students have to name for themselves.

DOI: https://doi.org/10.18061/ojsm.issue.389