The wealthy have rejected AI tutoring for generations, preferring traditional schools where children learn alongside peers. Even as tech companies pour billions into personalized learning systems, the pattern persists: Bill Gates sent his daughters to Seattle's Lakeside School rather than hiring world-class tutors, and Mark Zuckerberg created a small private school in his home for other children to attend. This disconnect between what tech advocates promise and what families actually choose reveals a fundamental truth about education that no algorithm has solved.

What Do Parents Actually Want From Education?

The gap between tech-enabled personalization and parental priorities is wider than most realize. According to Joe Liemandt of Alpha School, a company claiming its technology can double learning speed, only 10% of parents care about outstanding academic outcomes. The vast majority simply aren't motivated by the prospect of their eighth grader taking calculus early.

Instead, parents consistently prioritize the social dimensions of schooling. They want to know whether their children are taking school seriously, treating teachers with respect, and making friends. This social aspect of education, while sometimes promoting cultural homogeneity, remains deeply appealing to most families. The persistence of this preference across decades and income levels suggests it reflects something fundamental about how humans learn and develop.

Why Has Personalized Learning Failed to Gain Traction?

The dream of tech-enabled personalization isn't new. Skeptics point to a century of similar rhetoric, with quotes from 1920 sounding nearly identical to pitches from today's tech CEOs. The failures of platforms like RocketShip, School of One, AltSchool, Summit, and Khan Academy's Khanmigo to achieve their promised impact are often blamed on imperfect technology. But this explanation misses the real issue.

The fundamental problem isn't technical; it's that individualized learning has limited appeal, even in theory. Most people prefer social learning environments. This doesn't mean Silicon Valley won't eventually push personalized instruction into mainstream education, but if they do, it may leave many families unhappy with the outcome.

How Can AI Better Support Teachers Instead of Replacing Them?

A more promising direction for AI in education focuses on enhancing human teaching rather than replacing it. Researchers at the University of Massachusetts Amherst have received a National Science Foundation grant to develop AI systems that simulate realistic student behavior for teacher training.

The project addresses a core challenge: education is inherently messy and nonlinear. Real classrooms involve questions, errors, and sometimes arguments. Different students learn differently, making teacher preparation difficult with one-size-fits-all methods. Current AI models struggle to replicate this complexity.

• Simulated Student Agents: AI systems designed to act like real students in learning situations, helping tutors and teachers anticipate errors and improve engagement
• Interactive Learning Models: Unlike existing passive systems, these AI agents can hold conversations with teachers-in-training and respond to instruction in individualized ways
• Practical Applications: The technology will support tutor training, curriculum design, and research on optimizing AI for teaching effectiveness

"GenAI models, such as LLMs, are not good at mimicking real student learning behavior yet," said Andrew Lan, associate professor and the undergraduate computer science program director at UMass Amherst's Manning College of Information and Computer Sciences. "Existing student models are only designed for passive learning activities such as question answering; this project will create truly interactive student models that can hold conversations with teachers-in-training and engage with material in individualized ways, like real students."

Andrew Lan, Associate Professor and Undergraduate Computer Science Program Director, University of Massachusetts Amherst

The research team has three specific goals: reproduce realistic student errors, study how to infer engagement and understanding from open-ended learning activities like discussions, and evaluate how simulated students can improve educational approaches. Experimental studies with pre-service teachers will test effectiveness.

Lan emphasized that the primary focus is benefiting human tutors, not replacing them. "We are going to run some studies with real human teachers interacting with AI-powered teachable student agents," he explained. The technology could eventually support learning-by-teaching activities where students learn by explaining concepts to AI agents.

Lan

What Should Parents Know About AI and Their Children's Future?

As AI becomes more capable, parents face genuine uncertainty about how to prepare their children. Daniel Susskind, a researcher who has spent 15 years studying AI's impact on work and society, warns against simply banning or restricting AI use in schools. Instead, he argues for a more imaginative approach.

The traditional strategy of "future-proofing" skills no longer works. In 2013, England became the first country to mandate coding education for all primary and secondary students, with the goal of equipping children with 21st-century skills. By 2026, AI systems like ChatGPT and Claude became expert at writing code, with 90% of the code for Claude Code written by AI itself. Skills meant to protect children from technological disruption became largely redundant before they left school.

Rather than attempting to predict which skills will remain valuable, Susskind argues parents and educators should focus on foundational abilities. Since 2009, literacy and numeracy have declined among young people worldwide according to the OECD's Programme for International Student Assessment. These basics matter because all advanced skills, whether creativity or judgment, depend on them.

The challenge isn't choosing between embracing or rejecting AI, but preparing the next generation to flourish in a future that remains fundamentally uncertain. This requires imagination, flexibility, and a willingness to rethink education itself rather than simply adding technology to existing systems.