A quarter century ago, E.D. Hirsch Jr. published a cleverly titled essay: ‘You Can Always Look It Up’…Or Can You? His concern was prescient. The internet was placing information at our fingertips, but students weren’t building enough foundational knowledge to make meaning of it, or sometimes even know what to “look up.” Hirsch worried the advent of the internet would accelerate progressive education’s emphasis on “learning skills” at the expense of content knowledge, mistaking access to information for actual understanding.

This year, Dr. Barbara Oakley and colleagues released The Memory Paradox: Why Our Brains Need Knowledge in an Age of AI, providing neuroscientific confirmation that, indeed, you can’t “just look it up.” Cognitive offloading to AI prevents the brain from forming the mental schemas required for comprehension and true learning.

Yet the “Age of AI” also offers an opportunity for a knowledge revival, especially if we can bridge the divide between progressive, child-centric education and communal, knowledge-based education. Both sides care about equity, engagement, and preparing students for meaningful lives. Neuroscience and the risks of over-reliance on AI simply remind us that these aims require equipping all students with the foundational knowledge that makes learning possible. By “foundational knowledge,” I mean both the shared background and vocabulary all students need for common understanding and the domain-specific knowledge that enables deeper thinking within fields.

Why Knowledge Matters More Than Ever

As AI accelerates, addressing our “knowledge deficit,” as Hirsch called it, becomes even more urgent. A student who asks AI to explain climate policy can’t evaluate the response without foundational knowledge of science, economics, and civics. In fact, they likely don’t even know the right questions to ask. AI can synthesize information instantly, but it can’t give students the language and conceptual frameworks needed to comprehend and think critically about that information. Without knowledge in their minds, students become passive consumers (and “writers”) of AI-generated content rather than informed, engaged, and discerning users of a powerful tool.

What Knowledge-Rich Classrooms Actually Look Like

Some fear that teaching content means boring, rote, “industrial era” instruction. The reality is the opposite. In knowledge-rich classrooms, students are deeply engaged, not because content is tailored to their interests, but because facts and ideas, taught intentionally, cultivate curiosity and enable neural connections, pattern-recognition, and genuine thinking.

As one educator-parent shared on the History Matters podcast, she overheard her second-grader apply basic economics learned in social studies during an Easter candy trade with his older sister, confidently pronouncing, “Your supply is low and your demand is high, so it’ll cost you two Swedish Fish!” I’ve seen a fourth-grade class of predominantly immigrant students energetically discuss whether Columbus should be considered a hero or a villain, demonstrating the language and historical knowledge to grapple with complexity and a topic too often deemed by adults to be too controversial. I’ve observed a third-grade teacher bring her class back to attention by asking students to touch their patella, then their sternum, then their Achilles tendon – joyfully reinforcing vocabulary, literary allusion, and anatomical knowledge. And teachers implementing a knowledge-rich curriculum often share, “My students’ writing has improved dramatically because they finally have something to write about!”

This is what education can look like when we prioritize building knowledge. Students develop not just individual understanding, but the shared foundation that makes discourse, real-world application, and collaboration possible. This foundation also equips them to use AI and strengthens confidence, agency, and critical thinking – the very “skills” most progressive, child-centric educators seek to develop.

Two Types of Knowledge All Students Need

To prepare students for the AI era, all learners need two essential types of knowledge:

Shared background knowledge: The historical events, mathematical and scientific concepts, literary and cultural works, geographical understanding, and civic foundations that create common ground for communication and collective problem-solving. This isn’t about uniformity of thought. It’s about ensuring students have shared reference points that enable productive engagement across differences. When students study the Constitutional Convention together, they can debate modern governance challenges with shared context. Without that foundation, conversations fragment into isolated, ungrounded, and undebatable opinions.

Domain-specific knowledge: Deep understanding within disciplines that enables transfer and expertise. If shared knowledge enables communication, domain knowledge provides the disciplinary lens students need to internalize how scientists evaluate evidence, how historians interpret sources, how mathematicians model relationships, how engineers design solutions, and how writers construct meaning. As Charles Fadel argues in Education for the Age of AI, AI’s ability to replicate declarative and procedural knowledge means students must understand not just what a field knows, but how each discipline builds and evaluates knowledge.

Questions We Must Answer

Recognizing that shared background and disciplinary knowledge are the foundation, we must determine what that knowledge should include. As Fadel puts it, we must revise curriculum for the Age of AI “with a scalpel, not a chainsaw,” removing obsolete content, making space for modern disciplines, and ensuring teachers are not overwhelmed.

• What foundational knowledge do all students deserve? What’s truly essential versus merely traditional? How can we modernize the canon thoughtfully, ensuring the curriculum is inclusive and relevant while remaining manageable?
• How do we decide what to prune?  What procedures can be safely offloaded to AI without undermining deeper learning? Conversely, what content is indispensable for building the mental models students need to ask good questions, evaluate AI’s output, and build and transfer knowledge across domains?
• What domains and disciplines must we prioritize? How do we balance traditional subjects with emerging fields like data science essential for the modern world?
• How do we ensure coherence across the curriculum? How can we provide all students with access to a purposeful progression of shared background and domain-specific knowledge that builds year over year, aligns across disciplines and instructional materials, and enables deeper comprehension and discernment?

You Can Always Ask AI…But Should You?

Only if you know enough to make both the question and the answer meaningful – to engage in active learning rather than passive consumption. This is why foundational knowledge matters more in the Age of AI, not less.

Yet the stakes extend far beyond AI usage. We must accept this paradox: the more powerful AI becomes, the more essential foundational knowledge becomes for human flourishing. Answering these curricular questions with clarity, courage, and cognitive science as our guide will ensure students develop the ability to ask AI anything and the judgment to know when and how they should, but most importantly, the knowledge and understanding required to learn continuously, contribute meaningfully, and thrive as individuals and citizens.