Education is Failing our Students in this AI-Driven World

Education is struggling to adapt to AI as quickly as work and learning are changing, Chris Benson and Daniel Weitnach argue. Weitnach sees AI as affecting how people do their jobs more directly than social media, whose effects were concentrated in particular fields such as digital marketing and HR. Some schools initially told students not to use AI because they had not decided how to incorporate it into coursework. But students are likely to use increasingly capable tools throughout their working lives, so individual educators’ experiments need to be matched by changes to district and state curricula and requirements.

The challenge is not simply to teach students how to use AI. They also need enough subject knowledge to judge its output, ask useful questions, and work through problems with it. Basic math intuition, accurate communication, critical thinking, and creativity remain important even when AI can perform parts of a task. The speakers also caution against removing all difficulty from learning: struggle and discomfort can help students develop problem-solving, decision-making, and resilience. The goal is to reconsider which skills students need, not to treat either AI use or foundational learning as a substitute for the other.

Benson says AI has increased his output, but shifted where he spends effort. AI-generated code has made his Rust expertise less important to his work than it was a year ago, while giving him more room to focus on creative tasks. He argues that schools should identify which foundational skills AI is displacing and consider how those skills will matter in coming years. If education falls too far behind changes in work, he warns, today’s children will have a hard time. That does not mean every current skill is obsolete; it means educators need to examine what AI changes and what students still need to know.

One proposal is to publish the full year’s curriculum on the first day, leaving out tests and quizzes, so students can explore material early with AI and parents can devote more time to areas where their children struggle. Benson expects teachers to become facilitators of more customized learning, a role he believes would keep human educators highly needed rather than replace them. The speakers also argue that teachers and professors need more freedom to experiment. One suggested exercise is to have students critique AI-generated essays for inaccuracies and weak framing, using the tool’s mistakes as material for learning.

The research discussed suggests that results depend on how AI is used. A 2025 field experiment with nearly 1,000 high-school mathematics students in Turkey found that a general chatbot improved performance during assisted practice, but students’ subsequent unassisted exam performance was 17% lower than that of a control group. In a Harvard study of around 200 undergraduates learning physics, a carefully structured AI tutor produced stronger immediate learning gains across two lessons than an active-learning classroom comparison. The speakers stress that teaching practices vary by department and that these studies do not establish one best approach.

Purdue’s AI competency requirement across undergraduate degrees is cited as a positive, incremental effort to move education forward. The speakers contrast institutions that discourage or try to detect AI use in coursework with workplaces where comfort using AI in day-to-day work is expected; some professors may resist or police AI because they feel constrained or do not know how to respond. One speaker says a PhD developed confidence, critical thinking, problem-solving, creativity, and comfort with ambiguity, especially through producing novel research and defending ideas to skeptical experts. That process, rather than any single discovery, was the degree’s central value for that speaker, though entrepreneurship and creative work can also build these abilities and a PhD is not for everyone. Introducing more ambiguity, creativity, struggle, and tools earlier could preserve fundamentals while reducing reliance on memorization; education may continue to build these capacities, but through a different process.