Dear colleague,
Over the summer I came across this article from Barbara Oakley et al. Her team's findings are straightforward: even in our age of AI and instant information, ya still gotta know things. AI is like a souped up calculator — its usefulness is multiplied by the intelligence of its user, and that intelligence is built on a mountain of knowledge.
For longtime readers (e.g., Chapter 3 of These 6 Things), this is obvious stuff. But let's dig a bit into the science for a minute, as it's both fascinating and practical.
The Research
Oakley's team unpacks what we know about how the mind learns and remembers. They describe our two collaborative memory systems: one that consciously stores facts (like when we deliberately memorize our students' names), and another that makes skills automatic through practice (like how we eventually don't need to think about what our students' names are — we just know them).
Here's the key insight: when students constantly Google (or ChatGPT) information instead of learning it, they never make the shift from conscious recall to automatic expertise. They stay stuck in the slow, effortful mode of thinking because they never build the internal knowledge structures that make advanced thinking possible.
(Veritasium's Derek Muller makes this case beautifully in this must-listen keynote.)
Think about it this way — an elementary student who always used a calculator for basic math can pass basic math tests, but they never develop number sense. Down the line, they can't spot obvious errors or see patterns. Without that foundation, higher-level math becomes nearly impossible — and ask your math colleagues if you don't believe what I'm saying because I've never met one that can't speak to this reality.
So What?
Oakley and her team are validating what many of us have long known: memorization isn't the enemy of critical thinking — it's the foundation.
When I wrote about knowledge-building in These 6 Things, I argued that content knowledge is inseparable from skills. You can't think critically about something you know nothing about. After one article of the week on North Korea, my students want to nuke the country; after four articles of the week, they leave the nukes behind and start brainstorming humane, intelligent solutions. (That story's on pp. 72-73 and 81-84 of These 6 Things.) Oakley et al's paper demonstrates the neurological realities undergirding this.
Students need stored knowledge to:
- Recognize patterns and make connections
- Ask good questions
- Evaluate new information they encounter
- Think efficiently and creatively
Without that internal knowledge base, students are constantly overwhelmed by cognitive load. They're using all their mental energy just to keep track of basic information, leaving nothing left for deeper thinking.
Should we be surprised, then, when students aren't motivated to learn while in this constantly exhausted state?
And Then Comes AI…
Misunderstandings around these realities have been around since I started my career in the age of Google. “If they can just Google it, why learn it?”
With AI's unasked-for advent, we're seeing the same episode. It's a rerun! And so now it's even more important to dispel myths around the unimportance of knowing lots of stuff. (Kelly Gallagher and I talked about this when discussing his recent book To Read Stuff You Have to Know Stuff: Helping Students Build and Use Prior Knowledge; that conversation is recorded here.)
Oakley's team found that students who rely too heavily on AI for writing and problem-solving show weaker brain connectivity in regions linked to focus and memory. They develop what researchers call “metacognitive laziness” — they let the tool do the thinking instead of building their own cognitive muscles.
So, guess what kinds of students benefit most from AI? Those who already have strong foundational knowledge.
They can evaluate AI output, ask better questions, and integrate suggestions meaningfully. Students without that foundation just accept whatever the AI produces.
How Then Shall We Teach?
None of what I've said is an argument for drill-and-kill-style instruction. But it is an argument for bringing great intentionality and a Feast of Knowledge to our students in all classes. We can do this by:
- Having students memorize key information. Yes, multiplication tables. Yes, vocabulary words. Yes, historical dates (see pp. 117-118 in The Will to Learn for how I approach date memorization in my history classes) and scientific formulas. Frame this as building their internal knowledge bank — the foundation for everything else they'll learn.
- And pro tip: create experiences in which students use the memorized information in multiple ways so as to make it meaningful rather than rote. E.g., “Write a paragraph describing how 5 dates on our list of 20 memorized dates connect to each other.”
- Use spaced practice. Don't introduce information once and move on. Circle back repeatedly so knowledge moves from conscious recall to automatic retrieval.
Balance tools with brain work. Let students struggle with problems using their own knowledge first, then use external aids to check their thinking or explore further. - Use low-stakes quizzing. I give a detailed treatment of this in this blog article.
The Gist
Even in an age when information is everywhere, we live and think from what we know in our heads. That's not going away until Elon Musk convinces us to put Neuralink microchips in our skulls. (No, thank you.)
Knowledge is what allows us to make sense of all that external information, to ask the right questions, and to think clearly about complex problems. It gives us access to the cosmos and the breadth of human experience.
Knowledge-building is beautiful.
Our students need internal knowledge structures to navigate an information-rich world. That's not old-fashioned — that's essential preparation for their future.
So let's embrace the hard work of knowledge-building. Not because we're nostalgic for the past, but because neuroscience shows us it's exactly what our students' brains need to flourish.
Teaching right beside you,
DSJR
nicoleandmaggie says
This is a really good article! (And not just because it includes some conclusions I had already come to on my own, including the calculator/number sense analogy…)
Dave Stuart Jr. says
Thank you! Grateful to hear it resonated 🙂