Teaching Isn't a Decision Tree

In this episode we explore the rise of AI agents in education through the lens of Grammarly’s expanding vision for AI in the workplace and the classroom. We examine the idea that teachers might one day “build” agents that encode their expertise, then ask a harder question: what gets lost when teaching is reduced to a set of rules? We look at what this means for teacher expertise, student equity, and the future of classroom decision-making.Topics covered:What AI agents are, and how they differ from chatbotsGrammarly’s broader vision for agent-based work and learningWhy the idea of “scaling teacher expertise” is both appealing and problematicThe limits of rule-based systems in real classroom instructionWhy pedagogical content knowledge cannot be easily encoded into an agentWhat Teachers Pay Teachers can teach us about quality and scalability in EdTechWhy grammar instruction and history instruction are not the same kind of teaching problemWhy equity concerns should be central when schools evaluate AI agentsThree key questions educators and district leaders should ask before adopting agent-based toolsSources:https://www.theverge.com/podcast/898715/superhuman-grammarly-expert-review-shishir-mehrotra-interview-ai-impersonationhttps://www.gartner.com/en/newsroom/press-releases/2025-08-26-gartner-predicts-40-percent-of-enterprise-apps-will-feature-task-specific-ai-agents-by-2026-up-from-less-than-5-percent-in-2025