The first five moves were:

1. Value Process over Product

2. Prioritise Live Learning

3. Provide Clarity about When and How to Use AI

4. Eliminate Gotcha Culture around AI Use

5. Expect and Model AI Transparency

For more on these strategies catch up on Part 1 of Sleight of Mind.

These next five moves turn from the climate around the task, to the task itself. These are moves that protect the “Aha!” moment so that the learning can’t be snatched whether AI is in the room or not.

AI is built to make everything easy, but learning requires friction. Our minds need to be challenged to think in new ways in order to carry a skill or concept from one context into another. Let AI be productive in our place, and we can admire the result but we won’t have been changed by the experience. We won’t have learned. And the research keeps telling us that this kind of production doesn’t stick.

So how do we ensure students will have some opportunity to wrestle with a challenge whether AI is in the room or not? Here are five more moves to try this year.

Move 6: Reflect Before, During and After Learning

Reflection has a bad reputation among students because it’s usually bolted onto the end of an assignment. This reduces it to an irrelevant add-on, not the main event, and something to be rushed through in order to complete the assignment. Students write the paragraph you asked for, they tell you what they think you want to hear, most of it is descriptive (I did this and then I did that) and all of it is irrelevant to the current assignment because by then the work is already over.

Weave reflection through the process instead, and you open the door to change while there’s still something left to change. Three reflective opportunities to build in: a statement of intention before the work begins (What do I want to achieve with this? What am I curious about? What am I excited to create?), a halfway check-in (What’s working? What am I unsure of? How do I move this to the next level?), and a final reflection on how the student can transfer what they’ve learned to whatever comes next. In the age of AI, this is no longer a nice-to-have. It’s what keeps the student invested enough in the work to protect it, not just produce it

A new study by Ates found that students who evaluated their own writing first by naming what was working well and what still needed development, retained the ideas and arguments weeks later. However, students who used the draft stage as a brain dump and then skipped straight to AI feedback, improved faster in the moment, but that improvement evaporated the moment the AI wasn’t there to lean on.

The key takeaway is that not all AI feedback is equal. Students need to build out their ideas, wrestle with the initial creative mess, and reflect deeply on their own progress before AI gets a say.

Metacognition helps students frame what they want from an assignment, gauge how far they have gotten in achieving that ambition, and at the end, carry what they’ve learned to the next challenge.

Move 7: Put Students in the AI Driver’s Seat

The problem with how many students use AI is that the technology is built to do all the work and users slide into being passive accepters of whatever it hands them. That’s the road to offloading and deskilling. But teachers can train students to put the friction back in by building struggle into the task itself.

Building in an opportunity to struggle is a part of the Pedagogy of Productive Struggle, a framework developed by Lincoln Smith, Damion Walker and myself. The struggle verbs below are based on Bloom’s Taxonomy, adapted for AI interactions and designed to keep student agency. They can’t be used until a student has produced something to start with and the interaction that follows should be visible, shared with the teacher, and reflected on: did the AI help, or did it get in the way?

Once a student has built an initial draft of a project and they’ve reflected on what they’ve done well and what they need to improve, provide a menu of options for how they can engage with AI to iterate their work. Students should choose the provocation based on the next steps identified in their reflection.

Justify | Explain, defend, or critique the process, investigation, or deduction. “Ask the AI to disagree with your idea. Then explain why you still think you’re right or how the AI helped you change your mind.”

Modify | Revise, rethink, refine. “Ask the AI to identify weaknesses in your draft. Revise your work based on one of the suggestions, then explain what has improved.”

Classify | Evaluate, judge, critique, and prioritise. “Ask the AI to point out the strongest and weakest parts of your three ideas. Decide whether you agree, and explain which idea you think is actually best, and why.”

Unify | Find connections between disparate things. “Give the AI two contrasting ideas or points of view. Prompt it to propose three ways they might relate. Then build on its suggestions to create a fourth, original link.”

Complexify | Move beyond initial thinking; explore tension and ambiguity. “Ask the AI to find a tricky question or problem hidden inside your idea or argument. Use that challenge to make your thinking more interesting or detailed.”

Amplify | Imagine how to scale the idea up in a real-world context. “Ask the AI how your idea could become bigger or reach more people. Pick one suggestion and sketch what that scaled-up version could look like.”

The verb does the work. In each case the AI is still in the room, but the student is the one steering, and the thinking that matters stays firmly on their side of the desk.

Move 8: Condition Productive Failure

We know that student-driven discovery fuels learning. It’s the whole rationale behind inquiry-based approaches. We also know that when students reach for answers before they’ve built mastery, they often arrive at the wrong ones, which is why we lean on constructivist strategies to help them build meaning. But maybe what we need to do more of is design learning tasks so complex and so challenging that they will inevitably produce some wrong answers, so that being wrong is the point.

This is hard for us, because the Zone of Proximal Development tells us to provide enough support that a student can bridge the gap between what she can do alone and what she can do with help. The support part is our jam. We’re wired to engineer success, not failure.

In 2008, Manu Kapur argued that engaging students in complex, ill-structured problems without support could result in something he called productive failure. Struggling primes the mind to recognise why the correct method works once it’s finally taught. It isn’t an unfortunate side effect to be minimised. Struggle is the mechanism that makes the instruction that follows resonate.

Productive failure only works, though, if the student can tolerate the failure part long enough to reach the productive part, and that’s a Growth Mindset challenge. The two ideas need each other. So alongside our instinct to engineer success, we need to build in low-stakes room to fail and treat that failure as a milestone, not a mistake.

Move 9: Red Team Your Assessments

At the start of Part 1 of this two part post, I asked a question: could AI do the thinking that leads to the Aha? This move turns that one-off diagnostic into a habit.

Before a new tech product or model is rolled out to the public, it is often “red-teamed”. Red teaming is when an organization deliberately has a person or group try to attack, break, or exploit a system in order to find its vulnerabilities before real adversaries do.

Before you assign anything, instruct the AI to genuinely try to break your own assessment. Feed it your exact instructions. Hand it your rubric. Ask it for the fastest possible route to full marks. Don’t do the polite version of this; do the version a stressed, capable, time-poor student would do at eleven at night. If the AI hands you a credible product that skipped the struggle entirely, you’ve found the flaw before your students did, which is precisely when you want to find it.

Then fix it, and you already have the tools. Ground part of the task in something that happened in the classroom, so the AI can’t reach it (Move 2). Integrate an AI log so that its use is transparent (Move 5). Shift the weight of the grade onto the process steps rather than the polished product (Move 1). Red teaming is the quality-control pass that tells you which of the other moves this particular task actually needs. It doesn’t take long and it’s the difference between hoping a task is AI-resistant and knowing it.

Move 10: Create One Blue Sky Assessment a Year

Once a year, give your students a wicked problem: a real-world challenge, aligned with the concepts and skills you’ve built all year, that’s genuinely contested and unsolved: a problem where reasonable people disagree and there’s no single right answer waiting to be found.

Examples could include:

The cafeteria challenge: How do we reduce waste and cost while optimizing for taste and nutrition? Who are the actual clients of a cafeteria, and what is it for?

The phone policy: Should phones be banned, tiered, or contract-based by class? How do we balance focus and safety against the fact that a ban just starts a game of cat and mouse?

The AI-era school: Can AI teach? What does a modern school need to preserve, and what should it change?

Literacy, reading, and numeracy: Reading and math scores are in decline. What, if anything, should be done about it, and what value do these skills hold now and in an AI future?

There’s a second way to blue sky an assessment, and it doesn’t need a contested problem at all. Take a task you already run and blow the scope wide open. Instead of a short story, a book. Instead of a history essay, a series of podcasts. Instead of a lab report, a public exhibit. Instead of a persuasive essay, a real campaign aimed at a real audience.

The content and skills stay the same. What changes is scale, and scale is its own kind of friction. AI can write a competent five-paragraph essay in one pass; it cannot sustain a coherent book, or a dozen well-researched podcast episodes in a consistent voice, without a student doing the sustained planning and judgment-calling that holds a bigger project together.

Either way, the principle is the same: AI cannot steal a struggle it can’t shortcut in one pass. It can summarize, suggest, draft but it can’t hand a student the answer to a question that doesn’t have one, or hold sixty pages of judgment calls in its head on the student’s behalf. That work stays with the student, because there’s nowhere else for it to go.

A blue sky problem is a challenge to run and assess, and it only works once students have the foundational knowledge to engage with it seriously. It’s the reward at the end of a year of narrower, scaffolded work, not a replacement for it. But the effort is worth it. These types of assessments: ambitious, amorphous, student-driven and inquiry-driven are how students find meaning in school. How they can see the result of their learning efforts.

Eric Hudson put it well in a recent Learning on Purpose post: schools might do better modeling themselves on gyms rather than on workplaces because a gym is built on the belief that the unglamorous, repeated effort is the point, and the reward lives in the future. In the workplace, by contrast, no one much cares how you arrive at the goal as long as you deliver: promotions, raises, recognition, are all tied to hitting that mark. Swap “goals” for “grades” and you’ve got a pretty accurate description of the transactional school model that incentivizes unethical AI use.

As much as I agree with the analogy, it’s important to note that workplaces are great learning laboratories as well. Goals are set with no recipe for achieving them. Individuals and teams have to figure out how to use whatever resources, information, and tools they have. There’s real human effort in getting there, but also smart, resourceful use of people and technology.

The beauty of a blue sky assessment is that it is both gym and workplace at once: the one day a year you take the training wheels off and let them ride. It’s analogous to the piano recital, the playoff game, the product launch and the pitch that finally lands. It’s the moment that shows a student what all the earlier, narrower work was actually for.

Students have been finding ways around the hard part of learning for as long as there’s been a hard part to get around. AI just made the sleight of hand faster, smoother, and far easier to miss. This means we must design tasks worth struggling through, create conditions of trust and transparency, and be clear about our goals and methods. Then AI can’t snatch the struggle.

That’s the work ThinkSlant exists to do, and I’d love for you to do some of it with us. Slantwise isn’t going anywhere — I’ll keep thinking out loud here every week. But you’ll also find these essays cross-posted on the ThinkSlant blog, alongside classroom-ready tools, a lesson exchange, and a practitioner journal built by teachers, for teachers.

Sign up for the newsletter at thinkslant.org. And let’s make sure the struggle stays exactly where it belongs — with the students doing the learning.

The first is seasonal. For those of us in the Northern Hemisphere, these months bring some well-earned rest, but they also bring planning time for next year. In that spirit, this piece is meant to sit beside you while you do that planning. Over the next two slantwise posts I’ll offer ten teacher moves for building productive struggle into learning tasks. It’s more practical and less philosophical than what I usually write here, and that’s on purpose.

The second reason is one I’m really excited about. Together with some colleagues, I’m launching a platform for teacher collaboration as we navigate everything AI has brought into our classrooms. It’s called ThinkSlant. If slantwise is where I think out loud about all of this, ThinkSlant is where we try to build practical answers together. It’s a place to trade lessons that actually worked, to plan with tools designed for an AI-native classroom, and it’s a place to write up your thinking for a practitioner journal built by teachers, for teachers. This post is a sample of the kind of thing we’re making there. You can sign up for the newsletter at thinkslant.org if you’d like to follow along.

Opportunity for Productive Struggle

The trouble with AI – its whole seduction – is that it was built to make everything easy. Learning refuses to be easy. Learning is friction, effort, the productive ache of not-yet-knowing. Untrained and unguarded, students trust AI the way a mark trusts the crowd in a den of thieves: the struggle gets lifted, clean and quick, and what’s left in the empty pocket is cognitive offloading and deskilling.

Here’s how to catch the thief in the act. Choose a learning activity you’ve used before, something that asks students to do their own work. Now find the one crucial moment of cognitive struggle in it. The moment that usually produces the “Aha!” That moment is the prize. That’s what the AI is after. So ask yourself: could AI do the thinking that leads to that Aha? If the answer is yes, the learning is already at risk. The task needs to be redesigned. Build it so nothing can route around the friction. That friction is the point. Learning requires it.

The ten moves that follow are how we build productive struggle into learning and how we make the pocket impossible to pick whether or not AI ends up in the room.

The first five, the ones in this post, aren’t really about the task at all. They’re about the climate you build around it: the expectations, the vocabulary, the relationships, the habits students bring to the work before they’ve started the learning. Get the climate right and the tasks get easier to protect. In the next post, I’ll focus on the task itself, and about designing the kind of struggle that AI can’t do on a student’s behalf.

Move 1: Value Process over Product

This year, I asked a student why he’d chosen to open his essay with a counterargument rather than his own claim. He couldn’t tell me. He had an excellent product and no memory of having made a single decision to create it. I had little recourse other than trying to catch unauthorised AI use because my course outline valued the products and not the process. That’s what this move is built to eliminate.

Building process benchmarks scaffolds skills and knowledge and it gives us room to support the executive-functioning work students so often struggle with: planning and organisation, task initiation, time management, and metacognition. Since we have plenty of data suggesting that AI is simply too tempting to resist for homework, it matters more than ever to build out those process steps, because that’s where teachers actually get to witness learning happen.

It’s hard to argue with the value of scaffolding an assignment. But teachers keep running into the same two problems: there isn’t enough time for multiple benchmarks in every assessment, and students don’t take the formative steps seriously. So I’m going to suggest two things. Reduce the overall number of assessments, and treat the process steps as summative. You’ll end up with more time to assess and more assessment data, not less.

Imagine you’re a grade 11 History teacher. Over the year, your students typically face six major assessments: two tests and four projects or essays. That gives you six major marks, and from those you derive an overall grade.

Now reduce that to four assessments: two tests and two projects. Each project has four steps, and every step is marked. The steps vary in scope and challenge, but each one is keyed to a learning outcome. Suddenly you have ten data points to build a grade from. And I’d go further: if the final product is the last mark, the earlier steps should carry more cumulative weight than the product does. Remember, what we choose to assess tells our students what we care about.

Move 2: Prioritise Live Learning

One of the more interesting findings buried in the very popular MIT Media Lab study on AI’s cognitive impacts is this: unfettered access to AI may not lead to deskilling or offloading if that access is delayed until after students have first engaged in AI-free learning. It’s an early, promising result rather than settled fact, but it’s promising enough to build lessons around. And the AI-free engagement it points to isn’t just any engagement. Students need to have dug in, elbows deep, into the hard thinking of a genuinely challenging task first.

So how do we make that happen in a classroom, short of having everyone write an essay under timed and secured conditions? We prioritise challenging learning experiences that unfold in real time. This is the stuff of your observation lesson, the one where you pull out every trick to show engagement to the person writing your performance evaluation. Socratic seminar. Debate. Experiential learning. Think-aloud problem solving at the board. Peer teaching and jigsaw. The genuinely fun stuff. Any of these can stand on its own as an AI-proof assessment, but if you’re heading toward something students will eventually do at home, plant these in the early lessons that get them there.

Another approach is to anchor part of a written assessment in something that actually happened in class. For example, you might ask students to weigh competing theories by writing an analysis of a four corners exercise you ran, drawing on their classmates’ contributions to support their claims.

You can also build synchronous elements into the work itself: one-on-one check-ins, peer assessment or peer editing. And through all of it, mandate metacognition. Asking students to reflect on how, or whether, learning occurred, and to point to specific examples, does two things at once. It buffers against unauthorised AI use, and it helps students consolidate the skills and knowledge they’ve just built.

Move 3: Provide Clarity about When and How to Use AI

In 2024, when teachers were still reeling from the arrival of ChatGPT, Leon Furze and his colleagues gave us the clearest and most celebrated tool for integrating AI into assessment. Its latest iteration is more nuanced, and it takes a genuinely future-ready stance toward the kind of ambitious student-teacher collaboration that’s now possible.

There are three important ideas behind this tool. The first is that we can no longer set an assessment without deciding when, how, or whether students will use AI to complete it. Pretending it doesn’t exist isn’t an option. Even if you don’t want it used, you have to say so. Second, tasks have to be designed so that AI use, or its absence, is intentional. If no AI is to be used, how are you actually building that into the lesson? And third, most importantly to my mind, the scale gives everyone a common vocabulary which is essential to building clarity and trust between teachers and students.

Picture the student halfway through an assignment genuinely unsure whether she’s allowed to ask GenAI for feedback on her draft. According to a UK study, an overwhelming number of students report that they are frequently unclear on whether AI is or isn’t allowed for a given assessment. They’re asking for guidance on how to use it appropriately in their particular context, and they want to know how to avoid deskilling and offloading. So whether you reach for this AI Assessment Scale or another model, the imperative is the same: speak about AI often, and be clear about how, when, or if it’s allowed. And be explicit about what counts as good practice and isn’t. Every. Time. You. Assess.

Move 4: Eliminate Gotcha Culture around AI Use

I’ve written before about Gotcha Culture and the effect AI screening tools can have on students. Since then I’ve completed an Action Research Project for the International Boys’ School Coalition on using process-oriented strategies to mitigate AI dependency, and my thinking on how culture shapes AI use has deepened.

When we treat AI screeners as a game of catch me if you can, we risk two things: pushing unethical use underground, and, worse, reinforcing the idea that all AI use is wrong. If screeners are used at all, they have to sit inside a system of accountability guardrails — process benchmarks and the like — designed to open up conversation about AI use rather than shut it down. A screener flag can be one data point in the work of verifying learning, alongside an oral viva voce or a student’s rough notes. It cannot be the only measure of ethical AI use.

The screener’s fallibility can itself become interesting learning. Why are these tools sometimes wrong? How do they work? What kind of language gets flagged? Run a pre-AI-era paper through the screener as a class and see whether it trips. Run another and compare. Students come away understanding a good deal more about how the technology works. And if they’re preparing work for external assessment, say for the International Baccalaureate, it’s genuinely useful to run it through a screener beforehand to anticipate any questions about authenticity. That reframes the screener as a learning tool rather than a policing mechanism.

Long before Reichert and Hawley told us that positive relationships were the key to learning, we teachers already knew it. When students view their teacher as an ally rather than an adversary, they are far more likely to be honest about how they worked. A screener can flag a document. It cannot build the relationship that makes a student want to tell you the truth about how they wrote it. That relationship is the real accountability guardrail, and no piece of software has ever replaced it.

Move 5: Expect and Model AI Transparency

Students and teachers should talk about AI often: the good, the bad, and the ugly. And teachers should do this whatever their personal position happens to be, challenging themselves to be as objective as they can manage. Our students are walking into a world where this technology will touch everything. Simply telling them it’s bad doesn’t help them navigate that future ethically or productively. Embracing it without critical interrogation risks leaving them naive about its impacts.

Most importantly, we teachers need to demand of ourselves the same transparency we expect of our students. Depending on age, should be talking with our students about whether AI can or should provide feedback. Whether it’s acceptable to use for report cards. Or emails. We spend a lot of time telling students about what is off limits in their work when it comes to AI. Let’s invite them into conversations about our work. I think their ideas will be great for creating a set of principles to guide the class: what must be human created and can be co-created with AI, and why.

If we use AI to help plan our lessons, build our slides, or draft our feedback, we need to say so, in age-appropriate ways. I’ve lost count of the times I’ve attended teacher-training presentations where a slide carries an obviously AI-generated image with no citation at all. If you use AI to produce images, please hold yourself to the same standard that you expect of your students and cite! Even a tag at the bottom that says “This slide was co-created with AI” is a move towards a more transparent and judgement free culture.

We also need citation methods that are realistic and that actually reflect what AI is. I’m frustrated by how many AI policy documents insist that student AI use “must be cited” without reckoning with its ephemeral nature. That treats an AI exchange as though it were the same as a book on a shelf. Instead, students should keep AI logs, recording and reflecting on their use across a project: the prompts they tried, and an honest evaluation of whether the interaction actually helped. AI logs increase transparency and, almost as a side effect, ignite metacognition.

The Crowd, the Distraction, the Mark

Notice what all five of these moves have in common. They change the environment the task lives in: what we count, what we say out loud, whom students trust, and the habits they carry into the work before the work begins.

What does a pickpocket need? A crowd to disappear into, a mark whose attention is somewhere else, and a split second when the hand is already in the pocket. Take those away and there’s no lift. That’s what these five moves do. A classroom with real trust has no crowd to hide in. A classroom with real transparency offers no distraction to work behind. A classroom that pays attention to process leaves no mark looking the other way. And a thief who’s being watched, closely, the whole time, can’t work the con at all. None of this redesigns the friction. It just makes the friction much harder to slip past. Get the culture right and you’ve stripped AI of every condition its sleight of mind depends on.

In Part 2, we turn to the task itself, and to what it takes to build a struggle that AI genuinely cannot steal.

Not the silence of ignorance, or the blank look of someone who has nothing to say. This silence belongs to people who have put in real effort and believe they have reached a deeper understanding. And then someone asks them a question. The silence that meets it is the sound of someone thinking.

As a high school teacher, I’ve witnessed this silence many times and it is answered in one of two ways. The first is the gold standard of teaching and learning: the answer, when it comes, reveals conceptual depth and an appreciation for complexity. The student demonstrates mastery. The second has a slightly vertiginous quality, as if the student reaches for something they were certain was there and finds the shelf empty. The understanding they thought they had was, it turns out, more like a map of somewhere they'd never been. They read the map but didn't navigate the territory themselves, so they don't have much to say about it.

AI gives students the impression of mastery rather than deep learning. It provides the map so students think they don’t need to walk the terrain. Like many of my colleagues, I’m looking for ways to build productive effort back into learning, and I’m convinced that orality is one of them.

We recently played a game in my class. Students had just completed an outline for a 4,000-word research essay required as part of their IB Diploma. As a way to synthesise their ideas and hold the whole thing in their heads at once, they presented their paper to their peers in Just a Minute modelled on a BBC game show of the same name.

The trick of the game is that you present on a topic for 60 seconds without repeating, hesitating, or deviating. It’s far harder than it sounds. The students had 15 minutes to prepare. They could use their outline to rehearse, but had to do the full presentation without notes. Their classmates acted as judges.

It was a lively success, not only because 17-year-olds love a competition, particularly when they get to be judges, but because each student had to wrestle with representing a significant amount of work in a way that was accessible, interesting, and concise. They had to show what they knew through an authentic oral assessment. And in doing so, several of them discovered, some with obvious surprise, exactly where the shelf was empty. Where they didn’t walk the land and couldn’t remember the map.

Talking is not writing with your mouth. When you speak your understanding out loud, to another person, in real time, you discover what you actually know. Not what you can access with enough scrolling and searching. Not what you can assemble from good notes. What you have actually, genuinely, made your own. The question changes what you know. The act of reaching for an answer in front of someone, without a net, is itself a form of thinking.

This is old knowledge. Socrates used dialogue to teach. The medieval university assessed by oral disputation. The doctoral viva has persisted for centuries precisely because defending your thinking before another human being demonstrates mastery in a way no written product can replicate. What we have been doing for the last several decades — treating the written product as the primary, often the only, evidence of understanding — is the historical aberration.

AI forces us to reconsider that way of doing things, and to reconsider some older approaches. And perhaps some hybrid ones too.

The research on oral assessment is consistent: students who explain concepts aloud, who are questioned and must respond, demonstrate deeper retention and more flexible understanding than those assessed only on what they produce in writing. But I want to resist reducing oral assessment to merely an instrument for taking the temperature of learning that already happened.

What I find most compelling about oral assessment, done well, is that it is itself a form of learning. The student who stumbles on an explanation, who is asked “why?” and has to reach further, who corrects herself mid-sentence, that student is not just demonstrating understanding. She is deepening it.

Teachers working with large classes are already shaking their heads. A verification oral, a five-minute conversation after a written submission to check whether a student can explain what they apparently wrote, is useful: it addresses the immediate AI integrity problem. But it treats the oral as a secondary check on a primary product, and more importantly, it is almost impossible to resource. One-to-one conversations, five minutes per student, mean that a class of thirty or more would take days to assess. It also positions the teacher as auditor rather than interlocutor, which changes everything about what the conversation can do.

What I find more generative, and more capable of being deployed with large groups, are the formats where the conversation is the assessment, not an audit of something else: the seminar, the fishbowl, the Harkness discussion, the hot seat. Where the student’s job is to think out loud, in the presence of peers and a teacher. Where the teacher’s job is to listen carefully enough to ask the question that changes things.

Done thoughtfully, this can do something that written assessment almost never achieves: it can reveal the thinking of a student who could not otherwise express, or even realise, how deeply they understand.

None of this happens without explicit design. Oral skills need to be taught and scaffolded. Students need rubrics and time to practise at low stakes. It is also worth making an unequivocal statement that accent and grammar are not what is being assessed. For students who are anxious about public performance, who are working in a language that isn’t their first, or who come from classrooms where speaking up was risky, oral assessment done carelessly can disadvantage the same students that written assessment already disadvantages. Done well, however, the opposite can be true: students who struggle to express themselves in writing can flourish.

There is also the question of what oral assessment asks of teachers. Listening is a skill, and it is not the same skill as marking. The examiner who is waiting for a student to confirm the answer already in mind is not listening. The examiner who notices a hesitation, follows a thread, and asks “can you say that a different way?” is doing something genuinely demanding.

Just like the students, we need practice. Reviewing examples of other teachers doing it well — or having a colleague sit in and offer feedback — can build the skill considerably.

AI can help here too — from AI-generated Harkness prompts to scalable large-group oral assessments, educators are exploring hybrid approaches that support human-to-human dialogue and make classical oral assessment possible in the modern classroom.

The oral tradition in education is not a workaround for a new problem. It is a very old technology. And perhaps it is exactly what we need most right now — not because it outwits AI, but because it asks something of a student that AI cannot answer for them. It asks them to be present, in their own thinking, in real time. To walk the territory and not just read the map.

Toward a Community of Practice

I’m currently taking Leon Furze’s Practical AI for Curriculum Leaders course. As part of the course, I combined my recent interest in orality with AI-assisted coding to build a slide deck designed to support teacher professional development on oral assessment. I’m sharing it here in the hope that it may be useful to others exploring similar questions around assessment, pedagogy, and AI.

I’m also working on an exciting project that I hope to launch soon — a platform dedicated to connecting educators through the exchange of thoughtful ideas and innovative practices. This slide deck is the first contribution to what I hope will become a collaborative community committed to reimagining teaching and learning together. More on that to come.

Teacher PD for Oral Assessments

778KB ∙ PDF file

Download

Strategies, rubrics and research on using oral assessment to support student thinking in the age of AI.

Download

Wallace’s point wasn’t about fish. It was about the purpose of education which he argued was not the accumulation of knowledge or the credentialing of competence, but the cultivation of awareness. The ability to see what surrounds us so completely that many of us have stopped seeing it at all.

I’ve been returning to that image as AI seeps deeper into our classrooms, our assignments, and our students’ habits of mind. Because I think we are the fish. And AI is the water.

The Fish Don't Know They're Wet

The dominant educational response to AI has been AI Literacy: teach students to prompt effectively, to interrogate outputs, to understand bias and sycophancy and environmental cost. Teach them when AI helps and when it harms. Give them the framework, and trust them to use it wisely.

This is not wrong. I’ve built a version of it myself. But I’ve come to believe it isn’t enough, and that the gap between what it addresses and what it misses is exactly where the most important learning is happening, or failing to happen.

AI Literacy treats AI as a subject. What I think we need is something closer to what Wallace was pointing at: an education that changes how students understand themselves as learners, in a world that has been quietly, fundamentally, altered.

We’ve been here before, and we didn’t quite get it right.

When smartphones and social media arrived, we responded with Digital Literacy. We taught students to evaluate sources, spot misinformation, protect their privacy online. All necessary. But the most profound effect of that technology wasn’t what students were doing with it — it was what it was doing to them. Hana Lee Goldin recently wrote about this in Card Catalog. Attention spans shortened. Deep reading faltered. In the United States, reading and math scores among thirteen-year-olds peaked around 2012 and have declined steadily since.

We noticed. We lamented it. We assigned more books and wondered why students weren’t finishing them but we didn’t ask the harder question: what does it mean to teach reading — really teach it — to students whose entire relationship with text had been transformed? We kept handing them the same books, and they kept reaching for their phones, and we called it a motivation problem instead of a design problem.

I think we’re at risk of making the same mistake again.

Teaching the Swimmer, Not the Water

AI Literacy, even done well, tends to address the direct impacts of AI: the cheating, the deskilling, the uncritical acceptance of generated content. What it doesn’t yet grapple with is the indirect impact — the subtler erosion of a student’s belief that struggle is worth it, that their own thinking has value, that the crooked road of genuine learning leads somewhere worth going.

Eric Hudson, in Learning on Purpose argued that the arrival of AI makes our existing learning goals such as judgment, critical thinking, self-awareness, self-regulation, not less important, but more. The problem is that many of the tasks we’ve used to assess those goals are no longer reliable proxies for them. The task and the goal have come apart.

This seems to me exactly right. And it points toward what I’d call Learning Literacy, not teaching students about AI, but teaching them how to learn in a world shaped by it. Helping them understand their own minds: how attention works, how struggle produces growth, how motivation operates, how to notice when they’re reaching for AI out of fear or fatigue rather than genuine need.

In practice, this means designing differently from the start.

It starts with live learning which is a celebration of human learning: shared, experiential, and collaborative. No one is talked at. No one is passive. Students are already inside the work before they're ever asked to do it independently. By the time a student sits alone with a task and an AI tool, they are already elbow-deep in their own thinking. The ideas are theirs. The questions are theirs. AI enters a process already underway, rather than substituting for one that never began.

It means sequencing with motivation in mind. Not just asking whether a task is interesting, but asking why a student would care enough to do the thinking themselves. When the answer is unclear, AI becomes the path of least resistance. When the answer is genuine, students often don’t want AI to take it from them.

And it means designing tasks ambitious enough that AI’s presence changes the texture of the work rather than replacing it entirely. If a student is applying a theoretical framework from economics to a real problem in their school — the food in the cafeteria, the equity of the timetable — AI can support that work without doing it. The thinking that matters is still theirs to do.

Wallace was speaking to graduates in 2005, before the iPhone, before social media, long before any of this. What he was asking them to guard against was an unconscious absorption into the self and the blinkered assumption that the world exists to serve your immediate experience of it. Education, he believed, was the antidote: the slow, difficult cultivation of awareness of others, of context, of the water you’re swimming in.

That project feels more urgent now, not less. Because AI has made it extraordinarily easy to outsource the very thinking that builds that awareness. To skip the struggle. To arrive at an answer without ever having had the question live inside you long enough to do its work.

Teaching AI Literacy helps students use the tool. But the work Wallace was pointing at is protecting the human being doing the using. AI-Native education doesn’t ask students to resist AI or fear it. It supports them to remain themselves inside it: curious, capable of struggle, able to tell the difference between thinking they’ve done and thinking that’s been done for them. That’s not a small thing. In this moment, it might be everything. In fact, it’s the water we’re swimming in.

It’s free. It’s instant. It produces something that looks, on the surface, almost exactly like the real thing. And it has exposed something schools probably should have noticed sooner: we built our entire assessment system on the assumption that the currency — essays, lab reports, homework and every other learning product — was hard to fake.

It wasn’t, really. Students have always found ways to cut corners. But AI has made counterfeiting so frictionless, so available, and so good that the old economy can no longer hold. And the instinct of many schools to build better walls, mandate handwritten exams, and run everything through detection software, is understandable. It’s also the wrong response. You can make counterfeiting slightly harder. You cannot make it hard enough. Not anymore.

The real question isn’t how to protect the currency. It’s whether we should be running this economy at all.

Teaching Ethical AI Use is Not Enough

Here’s what makes this complicated. Most students aren’t confused about the ethics. They’re not looking for loopholes because they misunderstand the rules. After two years of teaching an AI literacy curriculum and surveying the students who came through it, what I found was not ignorance. It was something much more interesting.

They get it. Completely. They want to learn and they don’t want AI to steal learning opportunities from them.

I’ve repeatedly asked students what they’ve learned about using AI ethically in my course. Their responses could have been written by educators:

“AI is a tool we can use to learn. It is NOT a tool we use to learn for us.”

“Whenever I want AI to help me with an essay, I explicitly ban it from writing any essay snippets or ideas for me.”

“I will question myself when I am using AI to see if I am offloading work that is important to my development as a learner.”

They understand the risk AI poses to their own learning. They feel the pull toward authentic thinking. They can articulate, with impressive precision, exactly where the ethical lines are.

And then, at 11pm the night before a deadline, they cross those lines anyway.

This isn’t a knowledge problem. It’s a human one. And no amount of AI literacy curriculum, including mine, fully solves it.

The Knowing-Doing Gap

When I looked more carefully at the students who crossed their own ethical lines, I didn’t have to look far. This past year, a small handful of students used AI unethically on a final written assignment. I knew because the assignment had been built around process. There was a lesson that produced a short opening paragraph. A graphic planner. A conversation about the planner. A draft with feedback. And then the final product.

In two cases, the final product was unrecognisable from everything that had come before it. In the third, two of the middle steps were simply missing; the student had vanished from the process and reappeared only at the deadline, under pressure from home.

I met with each of them individually. No disciplinary hearing. No accusation. Just a conversation.

They all admitted it. Almost immediately. And what struck me most was that they seemed, in some strange way, relieved to have been caught.

The two who had completed the earlier steps weren’t lazy or indifferent. They were terrified. They didn’t believe that what they could produce on their own would be good enough. The process work was genuinely theirs but at the final moment, confidence failed them. The third had procrastinated himself into a corner so tight that AI felt like the only way out.

Two different problems. The same solution. One that was right there, frictionless, and free.

What I told all three of them was this: the consequence wasn’t a grade penalty. It was a learning penalty. By handing in work that wasn’t theirs, they had skipped the only part of the assignment that actually mattered, the part where their thinking got better. And because we had spent so much of the course talking about exactly that, they understood it. They owned the mistake. They redid the work.

That’s not a story about cheating. It’s a story about what happens when students don’t yet have the skills to manage pressure, doubt, and time and when a counterfeit printer is sitting right there on their desk.

Products Are Currency

Schools, if we’re honest about it, run like economies. Essays, lab reports, presentations — these are the currency students trade for grades, and grades are the currency they trade for futures. Learning, in this model, is the labour required to mint the coin.

AI has changed that completely. Not by making students lazier or less ethical, but by handing every single one of them a counterfeit printer. Free. Instant. Infinitely capable. And good enough that most people can’t tell the difference.

When we respond to this by building better walls by locking down browsers, mandating handwritten exams, running submissions through detection tools, we’re not solving the problem. We’re just making counterfeiting slightly harder. The economy itself remains intact. The incentive remains intact. And students, who are nothing if not resourceful, will find a way.

The uncomfortable truth is that schools created the conditions for this. We decided that what mattered was the product. We designed systems to assess it, rank it, and attach life consequences to it. We told students, in every structural way possible, that the destination was what counted. AI didn’t corrupt that system. It exposed it.

So the answer isn’t better detection. It isn’t more sophisticated assignments AI supposedly can’t do. The answer is to devalue the currency.

Process Is the Point

We need to shift the weight of assessment away from the product and onto the process: the incremental, visible, unglamorous work of actually learning something. Not as an add-on. Not as a soft alternative for students who struggle with tests. As the main event.

This costs something real. It means fewer assessments, because rich process takes time. It means stealing hours away from marking final products and reinvesting them in progress checks, drafts, reflections, and conversations. It means a course that used to include four major projects might only have two but those two will be genuinely witnessed, genuinely understood, and genuinely hard to fake.

It also means rethinking how we weight grades. If an essay has always been worth 20% of a term, consider awarding 15% to the process -- the planning artefact, the research benchmark, the draft, the reflection — and 5% to the final product. Not because the product doesn’t matter, but because it’s only one data point, and we’ve been treating it as the only one that counts.

And here’s what this gives back: when the process is the point, the counterfeit printer becomes almost useless. You can’t outsource the thinking if the thinking is what’s being assessed.

There’s something else it gives back too. When assessment is spread across the process, the stakes at any single moment become manageable. The student who’s running out of time has a progress check tomorrow, not a final deadline. The student who doubts herself has already handed in a plan, received feedback, and survived it. The conditions that drive students toward AI in desperation begin, slowly, to change.

This doesn’t solve everything. Students are human. They’re under pressure in ways that extend well beyond any single classroom or curriculum redesign. But it moves the conversation to the right place — away from policing what students produce, and toward supporting how they learn.

Learning as Assessment

My students already know how to be ethical. That’s what the data shows, and I believe it. They’re not waiting to be taught the rules. They’re waiting for a system that makes following the rules feel possible.

They aren’t bad kids making bad choices. They are capable people who ran out of time or ran out of faith in themselves, and reached for the easiest thing available. What made the difference wasn’t catching them. It was that the process work was already there, a record of their thinking, visible and real, so the conversation didn’t have to be adversarial. We could just look at it together and ask: what did you skip, and how do we go back for it?

That’s what process-centred assessment makes possible. Not just better detection of AI misuse, but better conditions for the honest conversation that has to follow.

The counterfeit printer isn’t going away. If anything, it will get better and harder to detect. We can keep building walls around it. Or we can ask what we were really trying to protect in the first place.

Learning isn’t a product. It never was. It’s the crooked road of struggle, mistake, revision, and gradual understanding. It’s the process we’ve been rushing past on the way to the grade. But you can’t outsource the thinking if the thinking is what’s being assessed.

Andy Hargreaves once warned that we risk valuing only what we can measure, rather than measuring what we value. AI has made that warning urgent in a way he perhaps didn’t anticipate. It has handed every student a way to produce the measurable thing without doing the valuable thing.

If we value thinking, we must give thinking value.

The reasoning is simple enough: if we don’t have some way of monitoring AI use, what’s to stop students from submitting AI-produced work as their own? That’s not a reckless or punitive position. It comes from a real concern for academic integrity and learning.

On the other side are those who argue that any use of AI monitoring tools, and the follow-up conversations they trigger, risk harming students because of false positives. Once a flag is raised, they argue, the interaction becomes accusatory. Monitoring tools are flawed, often biased against non-native speakers, and poor evidence on which to jeopardise trust. We are teachers, not police.

The truth, as usual, probably lives somewhere in between.

Anything to Declare?

I recently experienced my own version of AI policing, but this time I was on the receiving end.

After publishing my last slantwise article, The Human in the Loop, a reader ran a paragraph of my writing through Pangram, a new AI detection tool that claims extremely high accuracy. In the comments, they told me Pangram had attributed the paragraph to being written by AI.

I was horrified. Not because I don’t use AI, I do, but because I teach ethical AI use and hold myself to the same standards I ask of students. When I published the article, I felt confident in my process and fully believed the ideas, words, and structure were 100% mine. The accusation made me second-guess all of that. I felt guilty and defensive all at once.

The feeling was akin to being asked by a customs agent if I have anything to declare. Even when I’m fairly sure the $15 t-shirt I bought is well within the spending limit, the question alone makes me doubt what I know to be true. When authority asks, certainty wobbles.

It turned out the commenter was testing the screening tool as part of his own work as an educator and wanted to discuss what might have triggered the flag. It wasn’t a hostile comment, but it felt like it. And the emotional response lingered.

What stayed with me wasn’t the misunderstanding, it was the feeling it triggered. It’s humiliating. And even when you believe you’ve acted ethically, being asked if you did makes you doubt yourself.

And then I thought: this is how students must feel.

The Chilling Effect

Fear of being challenged about AI is having a chilling effect on students. One student I spoke with described her AI use as a careful, deliberate process—more restrained than many educators assume. She used AI for feedback, gave it explicit restrictions, and made sure it didn’t do the work for her. AI helped her plan, organise, and keep moving but she did the thinking and writing.

Even so, she didn’t declare her AI use.

What struck me most wasn’t how she used AI. It was why she wouldn’t disclose it. She was afraid that drawing attention to her AI use would trigger a suspicion of academic malpractice, not because she was hiding anything, but because “it’s not normal” to include a declaration and because she didn’t want to admit to doing something she hadn’t been explicitly permitted to do.

In other words, she was navigating invisible rules. She trusted her own integrity, but not the system’s response to transparency.

That’s the quiet harm of gotcha culture. Even ethical students learn that non-disclosure feels safer than openness.

Teachers Aren’t Okay Either

On the other side of the equation, teachers are feeling increasingly frustrated by ubiquitous AI use.

In my neck of the woods, the term is wrapping up and grades are being prepared. As final assignments trickle in, an alarming number of students are getting flagged by Turnitin for high AI use, use that isn’t cited or discussed with teachers. This even after students had taken AI literacy lessons where transparency was explicitly framed as essential to ethical use.

So how are teachers supposed to know whether AI was used ethically if students don’t feel safe being transparent?

A high AI flag has to trigger a follow-up conversation, but many teachers feel deeply uncomfortable initiating them. One colleague described it as an “adversarial approach to our responsibilities,” potentially compromising the relationships at the heart of good teaching. Others feel frustration, even betrayal, believing their students have crossed a line into cheating.

Almost all feel overwhelmed as yet another responsibility is layered onto teachers, further shifting accountability away from students and onto those assessing them.

What if everyone is right?

The Problem Isn’t AI. It’s Assessment.

I know I’m joining a chorus when I say this, but we have to change the way we assess.

In Five Principles for Rethinking Assessment with Gen AI, Leon Furze argues that reform shouldn’t start with policing AI use at all. It should start with good pedagogy.

1. Assessments must be valid and generate trustworthy evidence of learning. If they don’t, no amount of surveillance will fix that.

2. We need to design for reality. Students will encounter AI in real-world workflows. Pretending assessment exists in an AI-free bubble isn’t rigorous—it’s nostalgic.

3. Transparency and trust matter. Students want guidance. They want to do the right thing. But they can’t meet expectations that haven’t been clearly articulated.

4. Assessment must be understood as a process, not a moment. High-stakes, one-shot tasks are poorly suited to an AI-saturated world. Drafting, conferencing, reflection, and iteration aren’t add-ons anymore; they’re evidence.

5. And finally, we must respect professional judgement. Blanket rules and surveillance tools don’t just erode trust; they create unsustainable workload. Teacher expertise, exercised over time, still matters more than any detector ever will.

What We Can Do Now

These changes won’t happen overnight. Educational systems are notoriously rigid, and many schools are bound to external assessment frameworks like the IB or AP that limit radical redesign.

But there are changes we can make now.

We can prioritise live learning: experiential, inquiry-driven work that students are invested in. When students care about their thinking, they’re less likely to offload it. Assessments that require students to reflect on lived classroom experiences or build on them, remain harder to outsource to AI.

We can integrate learning portfolios across subjects, giving real status to reflections, drafts, artefacts, and oral defences of learning. Process evidence should stand alongside tests and timed tasks, not beneath them.

We can be explicit, again and again, about when AI is allowed, when it isn’t, and how to use it well for a specific task. Statements of AI use should be required for all assignments, so students get used to explaining how ideas, structure, and language were built, with or without AI.

Monitoring vs. Policing

None of this means abandoning accountability. Students are human. They’re under pressure. Sometimes they’ll make poor choices. Until assessment practices catch up, screening tools may still have a place, but only as conversation starters, not gotchas.

For that to work, teachers need training and support. School leaders must help teachers approach AI conversations with curiosity rather than accusation. I’ve seen this work. In a recent conversation with a student flagged for high AI use, I asked whether he noticed a pattern. He quickly identified that the flagged sections were those he felt least confident about. Once unpacked, he could see that while his overall process was ethical, he leaned more heavily on AI for language and ideation where his confidence dipped. Because we had rich process data, the conversation felt safe and productive.

The key is to shift the culture from policing, which is rule-focused, to monitoring, which is learning-focused.

All of this takes time. But the first step is moving away from product-heavy assessment. When assessment is designed well, when it’s valid, authentic, transparent, and process-oriented, an AI flag becomes just another data point for learning, not a verdict.

Until then, the real danger of gotcha culture isn’t that we’ll fail to catch misconduct. It’s that we’ll teach students that honesty is risky, trust is conditional, and learning is safer in in silence.

And that’s an assessment failure we can’t blame on AI.

But I find myself wondering: Who is the human we are keeping in that loop? How does that person react to confusion, friction, or doubt? How do they view themselves as thinkers in a world where production has become AI-turbocharged? How do they respond to temptation or delayed gratification? What pressures do they face in this new landscape?

When you shift the focus from the workflow to the person inside it, motivation suddenly becomes the whole story. Even when a student dutifully employs “the human in the loop” strategy when using AI, there is still ample room for cognitive offloading and deskilling. Motivation can be the key that determines whether or not this happens. Motivation isn’t a switch; it’s a narrative students use to make sense of struggle: “This is too hard,” or “I’m not good at this.” These stories don’t just influence behaviour: they sometimes give permission to avoid the discomfort that learning requires.

Students do not turn to AI for a single reason. Some are perfectionists, frightened that their own ideas won’t measure up. Some are impulsive, overwhelmed, or simply exhausted. Others are carrying far heavier things than schoolwork, and AI becomes a way of lightening the load. AI didn’t create these tendencies, but it meets them: smoothly, instantly, and without judgment. For the student who fears getting it wrong, AI offers polish. For the student who feels behind, it offers speed. For the student who is overwhelmed, it offers relief.

I’ve seen this play out repeatedly. Recently, the grade 12 students at my school were given time off timetable to work on a research project required for graduation. They had spent months with supervisors, revising drafts and refining ideas. This day was set aside for final polishing. Late in the afternoon, a student writing a math research paper began walking through hallways and common spaces, laptop open, asking every passing adult with any math background to review his work.

His supervisor had already told him the paper was strong and had offered some suggestions. He had worked diligently for months and used every support available. Yet in the final moments, doubt overwhelmed him and his anxiety crowded out his confidence. Eventually he made a series of changes to the paper based largely on ChatGPT’s revisions.

We’ve all encountered students like him: bright, capable, conscientious, and undone at the last minute by the fear that their best might not be good enough. How do tendencies like this make students especially vulnerable to using AI in ways that undermine learning. Not out of deceit, but out of self-protection. The work is not to tell students to “try harder” or to imply that using AI is a moral failing. Rather, it’s to help them recognize their own patterns with honesty and curiosity. To give them the language to notice when they’re leaning on AI out of fear, or fatigue, or the simple human urge to avoid discomfort.

This kind of self-discovery begins with simple metacognition. Naming the feelings that accompany challenge, such as frustration, self-doubt, and anxiety, can validate student experiences and foster an environment where transparency is the norm. This creates openings for conversations about how and why to make good choices when it comes to AI use.

Recently I’ve been reading 10 to 25 by David Yeager. He critiques the “neurobiological incompetence model” which is the idea that adolescents are half-formed humans, limited by immature brains. Instead, he argues that young people aren’t defined by poor self-control; they’re driven by different motivations than adults, namely social standing, respect, and belonging. When we understand this, the problem of student motivation looks less like a deficit and more like a mismatch. Students aren’t unmotivated or resistant to effort; they’re motivated differently.

Yeager proposes that the most effective way to support young people is to pair high expectations with genuine respect and support. His research offers practical strategies: be transparent; ask questions rather than issuing directives; offer “wise feedback” that conveys belief in a student’s ability; make purpose and belonging visible; and frame stress as something that can fuel growth rather than signal failure.

Building on Yeager’s work, one of the most powerful things we can do is frame productive struggle as a source of pride. When students wrestle with an idea and arrive at clarity, they feel the satisfaction of having earned their thinking and of joining a community of young people who take intellectual challenge seriously. Our job is to pair high expectations with steady support that communicates, “You can do this, and the struggle is the point.” By valuing difficulty, normalizing frustration, and offering feedback that highlights growth, we help students see themselves not as avoiders of hard things but as thinkers capable of meeting them.

One tool I’ve been experimenting with is the use of AI intention statements. At the start of a task, after I’ve clarified what forms of AI use are permitted, I ask students to write a brief statement of intention: How and why do you plan to use AI? At what point in your process? For what purpose? They can opt out of AI entirely or use any of the approved strategies. After completing the task, they add a declaration of their actual AI use as part of their citation practices.

The learning lies in the comparison. Where did their intentions hold? Where did they shift? What triggered that change: confusion, time pressure, perfectionism, impatience? These reflections help students anticipate their own patterns and develop the self-awareness needed to make thoughtful choices.

Keeping “the human in the loop” has never been about inserting a person into a process that involves AI. It’s about cultivating the kind of human who wants to be in the loop—who still feels the tug of curiosity, the satisfaction of an earned insight, the dignity of figuring something out. And that has never been more challenging, or more necessary, than it is now.

What made it difficult wasn’t just the emotion, or even finding the right memories—it was the responsibility of capturing who he was and what he meant, but doing it indirectly. To use words lightly and judiciously. Not to tell, but to show. To invite the gathered friends and family to join me in making meaning of a relationship, and sorrow, and a life gone too soon. This type of communication is effortful. Effort is the key that unlocks a shared truth. It’s active. It’s a creation made by both speaker and listener.

What I learned in writing that eulogy is something I now see everywhere—in writing, in learning, and increasingly in AI: that meaning isn’t delivered, it’s discovered.

In Tell all the truth but tell it slant —, Dickinson said that truth “must dazzle gradually.” She seems to be warning that unless we do some work in discovering truth, we can’t recognise it when it’s presented to us. Poetry is an oblique art—built on indirection and subtraction. A reader of poetry must engage, wrestle, associate, and reflect. Reading poetry demands effort. But the payoff, the reader’s “superb surprise” -- a jolting embodiment of the poem’s feelings, ideas, and sensory experiences -- is a moment of deeply recognised truth. True learning.

Unlike the effortful meaning making of poetry, most students see learning in school as transactional: effort is a cost, not a benefit, and grades are the reward. If you can skip the struggle and go straight to the outcome—you win. In fact, from their point of view, the purpose of schooling is to achieve an outcome. The learning itself is not the point. If you can leap straight to outcome while eliminating, or reducing, struggle, then you’ve cracked the cheat code of school.

We know intuitively that this doesn’t apply when we are developing physical skills. No one believes that they can be a concert pianist or an NHL hockey player without hours (years!) of practice. It requires making and learning from many mistakes, resilience, and active striving. In fact, you’d be hard pressed to be a good amateur in these pursuits without serious effort. In the realm of the physical, we know with certainty that a desired outcome is unachievable without struggle.

However, the same is not universally believed in the realm of thinking. A common perception is that those who appear to learn effortlessly are doing it best. In fact, if a student experiences too much cognitive effort, if there is too much friction or too much struggle, they think the problem is with them. This self-perception may negatively impact motivation, engagement and self-regulation. Most students prefer to listen passively to an entertaining lecturer rather than struggle with an active learning task. One study found that when it comes actual learning versus perceptions of learning, the introduction of struggle creates an inverse relationship. When there is no struggle, students feel like they are learning but are not, and when there is struggle, they feel like they aren’t learning but they are.

Students have a misconception that learning is best when it’s easy. An equally damaging perception is that while there might be some interesting diversions along the way, school must be endured until a desired outcome is delivered. Nothing reinforces this dynamic more than AI. Whenever a student butts up against a challenge, AI can provide the smoothest of solutions. A student can leap over the problem and land directly on an answer. AI hasn’t created this aversion to struggle but it has provided a free and simple tool to achieve it.

The idea that the purpose of learning is the straight line from school to career is an illusion. It skips the only part that matters — productive struggle. Deep learning is the crooked road of missteps, recalibration and grind. It’s the thinking equivalent of piano scales or hockey practice — complete with frustration and drudgery. In the AI age, students need to learn how to use the technology in ways that preserve struggle: to sharpen thinking, question assumptions, and reveal blind spots.

Left on their own, students instinctively use AI to smooth away productive difficulty, undermining deeper learning. They will do this even when they believe they are using the technology ethically. Some preliminary research even suggests that when using AI in ways that reduce or eliminate struggle, students may be risking cognitive decline. The comparison to physical effort seems to hold here: you gotta use it or lose it.

I have been thinking a lot about this, and along with some colleagues, we have come up with some ideas and a framework for how to find the right level learning challenge even when AI is involved. I’ll be exploring those in upcoming posts.

Truth doesn’t arrive fully formed. Effort reveals it. Learning is like that. So is writing. So is this moment of AI. Slantwise will be an invitation to keep wrestling, to resist easy answers, and to embrace the work that helps us to see clearly—even if we must look slant to do it.

Tell all the truth but tell it slant — 
Success in Circuit lies 
Too bright for our infirm Delight
The Truth’s superb surprise 
As Lightning to the Children eased 
With explanation kind 
The Truth must dazzle gradually 
Or every man be blind —
                Emily Dickinson

The name slantwise carries two meanings that matter to me.

First, it comes from the belief that truth is not something handed to us clean and whole. Understanding is constructed, not delivered, and it often arrives indirectly. The best thinking requires angles, pauses, revision, and reflection. It takes time. It takes struggle. It takes patience with complexity.

Second, slantwise is a stance, a deliberate way of resisting the speed and flattening of thought that now surrounds us—especially with the rise of artificial intelligence. AI can generate answers instantly. But answers are not insight. Output is not understanding. Thinking is not a task to be automated. To protect the quality of our learning, we need to build habits of metacognition, ethical inquiry, reflection, and intellectual honesty. We have to build friction. That takes intention. That takes purpose.

What this space is

This newsletter is a place to explore:

• How we learn—and how we might deepen learning with AI

• How AI is reshaping learning and what we must protect along the way

• How writing clarifies thought and deepens understanding

• What reflection and metacognition look like in real learning

• How to preserve agency, curiosity, and meaning in a technological world

I’m not interested in predicting the future of AI or endorsing particular tools. I’m interested in wisdom—in understanding what it means to stay fully human while learning, teaching, leading, and creating in a world of machines.

If you read this newsletter, expect ideas explored honestly, arguments developed carefully, and complexity treated with respect. Sometimes I’ll offer frameworks. Sometimes I’ll share stories. Sometimes I’ll wrestle with unfinished thinking. Always, I’ll try to move deliberatively enough to think well.

About me

I am an educator and writer who has spent my career working with young people and teachers to build meaningful learning—learning that doesn’t just transmit information, but shapes character, agency, and thought. My work lives at the intersection of pedagogy, ethics, writing, and emerging technology. I speak often with students and teachers about how to use AI as a tool for thinking rather than a shortcut around it.

I believe in intellectual courage. I believe curiosity is a skill. I believe writing is a form of thinking. And I believe learning is a lifelong practice—something we craft with intention.

What I do

• Write about AI, learning, writing, ethics, and reflection

• Develop practical approaches to thoughtful, learning-centered AI use

• Teach and mentor students and educators

• Explore how we can think—not faster, but better

If any of that resonates, I hope you’ll subscribe. I don’t have final answers, but I have good questions—and I’d like to explore them slantwise with you.