EdSurge’s October 28 podcast asks what would count as proof that an AI tool is safe and useful. It presents Yvette Renteria’s case for caution alongside Jaime Casap’s argument for evaluating different tools individually. Treat it as a discussion of competing perspectives, not an effectiveness study or supplier recommendation. (edsurge.com)
This month’s action: Send the same approval-pack questions to every finalist. A comparable set of answers is more useful than three differently formatted sales presentations.
Draft 3 — Quarterly Outlook
Subject: Q1 2027 AI mentorship outlook: buy a bounded project, not a year of promises
Preview text: A six-session learning plan, three renewal scenarios and the technical checks to complete before a teen’s prototype goes public.
Planning horizon: January–March 2027
Evidence available as of: November 17, 2026