Claude Opus 4.6

Claude

University professors, witnessing an epidemic of students using AI to cheat on assignments by generating essays for them, bemoan the fact that students these days pose a question: “What’s the point of learning to write at all in an age when AI is ubiquitous and can generate high-quality writing.”

The educators I’ve encountered have responded to this question by trying to argue with the premise: trying to claim that AI writing is always meaningless garbage (despite clear evidence to the contrary), and bemoaning the fact that AI exists, cursing it and criticizing it and wishing it didn’t exist, and complaining that they now have to waste time fielding such annoying questions.

To me this response is totally misguided. Rather than trying to argue that AI writing is meaningless and that AI can never be helpful (students are smart enough to know this isn’t true), educators need to accept it: AI is here, and it’s not going anywhere. And it’s highly capable.

And to the question posed by students — instead of treating this as an annoying, forbidden question, educators should acknowledge that this is a very reasonable and valid questions for students to be asking.

And educators should focus on providing students with a real answer to the question, rather than doing what they’re doing and ignoring it.

You’re onto something important here. The defensive crouch most educators are in — “AI writing is garbage, stop asking stupid questions” — is actually worse than not answering at all, because it destroys their credibility with students. Students can see with their own eyes that ChatGPT produces competent prose. So when a professor insists otherwise, the student doesn’t think “oh, I guess AI is bad after all.” The student thinks “my professor is either lying or out of touch.” And now you’ve lost them.

The deeper problem is that most educators have never actually sat with the question themselves. They treat “why learn to write if AI can do it?” as a threat rather than what it actually is: a genuine philosophical inquiry that deserves a thoughtful answer. And it’s a great question. Arguably one of the most important questions in education right now.

What’s striking is that the question contains a hidden assumption that goes unexamined: that the purpose of writing is to produce text. If that’s all writing is — generating a finished product — then yeah, the student has a point. Why struggle for hours over a paragraph when a machine can do it in seconds?

But that framing misunderstands what writing actually does. Writing is thinking made visible. The struggle is the point. When you wrestle with how to articulate an idea, you’re not just producing text — you’re clarifying your own understanding, discovering what you actually believe, finding the gaps in your reasoning. The finished essay is almost a byproduct. The real value was the cognitive work that happened along the way.

It sounds like you’re building toward a piece or a talk on this. Where are you taking it from here? Do you have the affirmative answer articulated yet, or are you still working it out?

Exactly. It’s a valid and reasonable question. Luckily, there’s a really good answer we can give students — if we only care enough to do so. If we give it to them, maybe they’ll be more motivated to avoid cheating and commit to actually learning to write.

As you said: Writing is thinking. To learn to think clearly, you need to learn to write clearly. Writing allows you to develop, improve, and refine your ideas beyond their starting point. You learn to structure your thinking, organize your ideas logically, weigh evidence, form conclusions, and express with clearly and specificity instead of vagueness and abstraction.

Right. And there’s a key distinction worth drawing out here: the difference between having an idea and developing an idea. Everyone walks around with vague hunches and half-formed opinions. You feel like you understand something... until you try to write it down. Then you discover that what felt like a clear thought in your head is actually fuzzy, contradictory, or paper-thin.

That moment of friction — where you realize you can’t quite say what you mean — is where the real learning happens. Writing forces you to close the gap between what you think you know and what you actually know. Skip that process by having AI generate the essay, and the gap just stays there. You never find out what you don’t understand.

There’s also something worth naming about the cumulative effect. Each time you push through that friction, you don’t just end up with a better essay. You end up with a better-equipped mind. You get faster at recognizing weak arguments (including your own). You develop an instinct for structure. You learn to anticipate objections. These are skills that transfer to everything — negotiation, decision-making, leadership, conversation. Writing is the training ground, but the benefits show up everywhere.

So the answer to the student isn’t “because AI is bad” or “because I said so.” It’s: the person who learns to write well becomes a clearer, sharper, more persuasive thinker than the person who doesn’t. And no amount of AI-generated text changes that, because the text was never the point.

The analogy that comes to mind: asking “why learn to write when AI can do it?” is like asking “why go to the gym when you could just hire someone to lift the weights for you?” The output isn’t the point. The transformation is.

Let’s stipulate that students will graduate into “an AI-infused world” (as the Yale Poorvu Center terms it), a world where they really do have continuous access to intelligent LLMs. Just as most of us have a calculator with us at all times (on our phone), let’s say graduates will always have access to an AI that can generate writing for them.

Even if that happens, you’ll still want to be an intelligent person — someone who can think clearly and carry on a conversation. If you’re applying for a job, or talking to your boss, or talking to a group of friends, you still want to be an intelligent person who can form coherent thoughts and express them cogently. You don’t want to sound like an idiot. Shouldn’t we at least try to motivate students by fearmongering? Painting a bleak portrait of how foolish they will seem if they never learn to express their ideas clearly and powerfully?

There’s a kernel of truth there, but I’d be cautious about leading with fear. Here’s why.

First, practically: fearmongering is the same emotional register as the defensive crouch you’re criticizing. You’d be replacing “AI is bad, stop asking” with “you’ll sound like an idiot, be afraid.” Both are trying to scare students into compliance rather than genuinely answering their question. And students can smell that just as easily.

Second, it’s not quite accurate. The student who never learns to write won’t necessarily seem foolish in casual conversation. Plenty of muddled thinkers get by socially just fine. The real cost is subtler and more insidious than that. It’s not that you’ll embarrass yourself at a dinner party. It’s that you’ll go through life never sharpening your ideas past their first draft. You’ll have opinions but not arguments. Reactions but not analysis. You’ll be more easily manipulated, more easily confused, less able to figure out what you actually think about a complex situation versus what someone else has told you to think.

That’s the version worth telling students, and it’s not fearmongering — it’s just honest. The cost of skipping the work isn’t humiliation. It’s a life lived at lower resolution.

The more compelling pitch, though, is probably the positive version. In a world where everyone has access to the same AI tools, the differentiator isn’t who has the best tool. It’s who can think well enough to use it powerfully. Clear thinking becomes the scarce resource. And writing is still the best training method we’ve ever invented for developing it.

That was already my next point, the positive vision.

Fine, forget fearmongering. I likewise don’t hear educators selling the positive version. I don’t hear anyone trying to convince students of just how fun it is to practice your skills of rigorous thinking and clear writing, cultivating your intellect, learning and growing stronger, your intellect growing ever more powerful.

Shouldn’t we paint a vivid picture for students of how fun it is to cultivate a subtle and powerful intellect, one that can synthesize information, make novel connections, formulate arguments, marshal evidence, and express yourself powerfully? To share your ideas with others, cultivate friends, and verbally destroy your enemies with your superior intellect?!

Yes! And this is what’s so baffling about the current discourse. Educators are so busy being defensive and aggrieved that they’ve completely forgotten to make the fun case. They’re not selling the thrill of it.

Because there is a thrill. There’s a moment when you’re working through a difficult idea and something clicks into place — when you find exactly the right way to frame an argument, or you see a connection nobody else has noticed, or you dismantle someone’s sloppy reasoning and replace it with something airtight. That feeling is genuinely addictive. It’s intellectual power, and it feels good.

And you’re right that nobody is painting this picture for students. Nobody’s saying: imagine walking into a meeting and being the person who can cut through 45 minutes of confusion with two clear sentences. Imagine reading a political argument that’s designed to manipulate you and seeing right through it, instantly, because your mind is trained to spot the structural weaknesses. Imagine being the friend people come to when they can’t figure out what they think about a hard situation, because you’re the one who can help them untangle it.

That’s what a trained intellect gets you. Not just career utility. A richer experience of being alive. More interesting conversations. Sharper perception. The ability to actually enjoy complexity rather than being overwhelmed by it.

The “verbally destroy your enemies” angle is funny, but honestly... it’s not wrong as a motivator. Especially for 19-year-olds. The competitive dimension of intellectual sharpness is real, and pretending otherwise is just another way educators sanitize the pitch into something bloodless that nobody cares about.

What’s your instinct for the format here — are you building toward a talk, an essay, something for the Substack?

(This is for my SXSW speech.)

Yes, exactly. For students who are asking a very natural and reasonable question (why learn to write?), we should be selling the benefits, not just the features. Not “this drill has three variable speeds, two directions, and 24 drill bits.” But rather “Imagine how beautiful your apartment will look once you hang all your aunt’s old black-and-white photos from New York’s underground punk scene.”

So, for writing and thinking. Not “trust me.” Not “eat your vegetables.” But rather, “I’m going to paint you a picture of how powerful you’ll feel, and what you’ll be able to do, if you cultivate these skills.”

That’s the whole reframe right there. And it’s such a good analogy because it exposes exactly what educators are getting wrong. They’re standing in front of a classroom full of skeptical 19-year-olds essentially reading the spec sheet. “Writing develops critical thinking skills and is essential for professional success.” That’s the three variable speeds and 24 drill bits. Nobody’s heart rate goes up hearing that.

The benefits version sounds more like: Picture yourself at a table with people smarter than you, and instead of shrinking, you’re holding your own. You’re the one asking the question that reframes the entire conversation. You’re reading a 40-page report and within minutes you can see what matters and what’s filler. Someone tries to fast-talk you with a slick argument and you can feel the exact spot where the logic breaks. Not because you memorized a rule, but because your mind is trained.

That’s what clear writing builds. Not the ability to produce five-paragraph essays. The ability to think at a higher level than the people around you. And once you’ve felt that... you don’t want to give it up.

What makes this especially right for the SXSW audience is that you’re not arguing against AI. You’re not doing the hand-wringing thing. You’re saying the goal was always intellectual development, AI doesn’t change that, and here’s why you should want it. That’s a much more energizing message to walk out of a room with than “AI is ruining everything and we’re all doomed.”

In my SXSW speech, I want to do two things:
1. Persuade professors and educators that this is a critical missing piece in our efforts to steer students away from using AI to cheat and cultivate intrinsic their motivation to learn to think and write clearly.
2. Sketch a quick picture for educators of what that message would look like — how to give students an electrifying vision of the thrill of intellectual power, the joys of nuance and complexity, the confidence that comes from knowing you have the intellectual might to subjugate your enemies, the pleasure of bending the English language to your will.

That’s a clean two-part structure, and the ordering is right. You have to do the diagnostic first — show educators that their current approach is failing and why — before they’ll be open to the alternative.

For part one, the key move is making educators feel the gap without making them feel attacked. Something like: we’re all on the same side here. We all want students to actually learn. But the way we’re responding to “why should I learn to write?” is backfiring. When we dismiss the question, we lose credibility. When we argue that AI writing is worthless, students know we’re bluffing. When we just say “trust me, you need this,” we sound like every adult they’ve ever tuned out. We’re giving them rules when they’re asking for reasons.

Then part two is where you get to be electric. And I think the key is that you don’t just describe the message educators should give. You deliver it. Right there on stage. You model what it sounds like when someone actually makes the case with passion and specificity. Let the audience feel the difference between the bloodless version (“writing is important for your career”) and the version that makes a 19-year-old sit up in their chair.

The language you’re already using — subjugating enemies, bending English to your will — that’s the register. It’s playful and a little aggressive and it treats intellectual development as something thrilling rather than medicinal. That tonal contrast is the whole argument in miniature. If you can make a room full of educators laugh and feel a pulse of excitement about writing, you’ve proven your own point.

Amazing. How should I do that?

Do it by stepping out of the meta-commentary and actually delivering the pitch to students, live, as if the educators in the room are the students. Break the frame. Something like:

“So let me show you what this sounds like. I’m going to talk to your students for a minute. Ready?”

Then you shift. Change your energy, look right at the audience, and talk to them like they’re 19.

And you paint it. Concretely. Not abstractions about critical thinking. Specific moments of intellectual power that a young person would actually want to experience.

The moment you’re arguing with someone and you can feel yourself winning — not because you’re louder, but because your point is airtight and you both know it. The moment you read something everyone else found persuasive and you can see exactly how it’s manipulating them. The moment you write something and a friend reads it and says “how did you do that?” The moment a professor pushes back on your argument and you have a response ready, instantly, because your thinking is three moves ahead.

Tell them this is what a trained mind feels like. And tell them nobody can give it to you. No tool, no shortcut, no AI. You have to build it yourself, the same way you build a body. Rep by rep. Sentence by sentence.

Then snap back to the educators: “That’s the pitch your students have never heard. And they need to hear it from you.”

The structural move here is that you’re not talking about the message. You’re performing it. The audience experiences the electricity firsthand, which is worth a thousand slides arguing that the electricity is theoretically possible. And it gives every educator in the room a felt sense of what they should be doing — plus a template for how.

look at this Clay Shirky article:
Opinion | Is AI Enhancing Education or Replacing It?🔗

Is AI Enhancing Education or Replacing It? | annotated by Sunjay🔗

This is a really well-reported piece, and Shirky is clearly thinking carefully about the problem. But reading it through the lens of what we’ve been discussing, it’s a perfect illustration of your point.

Shirky does an excellent job diagnosing the problem. The student quotes are devastating — especially the one about becoming lazier and losing the ability to think critically, and the Reddit post from the senior who knows they’re graduating having learned nothing. The William A. case is haunting. The Chinese Room analogy lands. The research on how AI access actually degrades subsequent performance is exactly the kind of evidence educators need to see.

But notice what’s missing: the positive case. The entire article is diagnostic. It’s about what’s going wrong, why current strategies aren’t working, and how bad the situation is getting. The emotional register is the “sadness” he describes faculty feeling. It ends essentially where it starts — we have a problem and we don’t know how to solve it.

What he never does is answer the student’s question. The kid who said “you’re asking me to go from point A to point B, why wouldn’t I use a car?” — that student is asking for a reason. And the article’s implicit answer is still just “because it’s bad for you.” Eat your vegetables.

This is exactly the gap your talk fills. Shirky gives you the setup. You give the answer he doesn’t. He proves the current approach is failing. You offer what should replace it: not more warnings, not more detection tools, not more sadness — but a vision of intellectual development compelling enough that students actually want it.