Shael Polakow-Suransky, president of Bank Street College of Education, witnessed a troubling scene in a Bronx seventh-grade classroom: students silently working with AI tutors on fraction problems while missing fundamental conceptual understanding. The technology corrected answers but failed to identify learning gaps, leaving both students and teachers unable to address why children struggled.

The essay argues that replacing human interaction with isolated screen time undermines deep learning. Students need productive struggle and peer dialogue to build lasting knowledge, not just procedural skill drills. This concern has sparked organized resistance, with New York City parents demanding a moratorium on AI in schools and a national coalition of 250+ child development experts calling for a five-year pause on generative AI in K-12 classrooms.

Current educational technology trends favor one-to-one AI tutoring, but this approach risks creating new forms of segregation. Wealthy districts may preserve human-centered learning while under-resourced schools—often serving Black and Latino students—receive laptop-based instruction. The core issue: AI can memorize formulas but cannot teach ethical debate or foster the collaborative skills students need for complex problem-solving.

Some educators see potential for AI tools that support project-based learning rather than replace it. Platforms like Playlab already host teacher-designed applications, while New York's new graduate framework emphasizes critical thinking over standardized testing. The real question is whether AI can measure meaningful learning outcomes instead of just counting easy-to-quantify responses.