New Year’s, And Preparing Minds For A World We Cannot Yet See

December 31, 2025

I've been noticing something in my conversations with students, usually late in our sessions when the guard has come down and the practiced answers have been exhausted. I'll ask them to explain their reasoning on some problem we've been working through, perhaps a complex passage of rhetoric, or a mathematical proof, or the underlying structure of an argument, and they'll pause. Then, almost apologetically: "Would it be cheating if I asked ChatGPT?"

The question itself is less interesting than what it reveals. Not whether such tools should be permitted. That's a policy matter, endlessly debatable. But rather what it means that these young minds, sharp and capable, have already begun to wonder which kinds of thinking belong to them, and which can be outsourced. Where, exactly, does their intelligence end, and the machine's begin? And more troublingly: have we taught them to tell the difference?

The beloved Columbia University professor Barbara Stoler Miller used to say that education's purpose was to make students "ungovernable by falsehood." She meant it quite literally: that a truly educated mind could not be easily deceived, manipulated, or led astray by sophisticated-sounding nonsense. I think of her words often now, as we watch these extraordinarily fluent AI systems produce text that sounds authoritative, coherent, even insightful, while also being, in ways that require genuine discernment to detect, subtly or profoundly wrong.

What strikes me about the current discourse around AI and education is how quickly it has collapsed into binaries: the enthusiasts who see liberation from drudgery, the skeptics who see the death of learning itself. Both camps seem to miss what's actually happening in the minds of the young people who will inherit this transformation.

The students I work with aren't asking whether they should use AI. They're already using it, with varying degrees of sophistication and ethical clarity. They're using it to brainstorm essay ideas, to check their code, to summarize dense texts, to generate practice problems. Some use it as a crutch, others as a catalyst. The technology itself doesn't determine which. What determines which is something we're not talking about nearly enough: whether they understand what thinking actually is.

Consider what colleges claim to teach: critical thinking, analytical reasoning, clear communication, creative problem-solving. These have always been somewhat nebulous goals, more aspirational than measurable. But now we have machines that can perform convincing simulacra of all of them. An AI can analyze a poem, construct a logical argument, write a persuasive essay, even generate novel solutions to certain kinds of problems. 

But it cannot, at least not yet, understand what any of it means. It cannot want to know something. It cannot be changed by what it learns.

This distinction matters enormously. However, it's not one we've trained our educational systems to cultivate. For decades, we've optimized for outputs that are now easily automated: the five-paragraph essay, the problem set completed correctly, the research paper with proper citations. We've built elaborate structures around measuring these outputs, credentialing them, sorting students by their facility with them. And now that facility, at least the surface performance of it, can be purchased for twenty dollars a month.

So, what becomes of the college-bound student who can prompt-engineer their way to an A but hasn't developed the deeper faculties that the A was supposed to represent? What becomes of the credential itself?

I'm reminded of what happened to penmanship. For centuries, beautiful handwriting was the mark of an educated person, a skill laboriously cultivated, a signal of discipline and refinement. Then typewriters arrived, and within a generation, the skill became quaint. We didn't stop communicating; we just stopped confusing the medium with the message. But here's what's different: handwriting was always just a transmission mechanism. The thinking was always separate from the hand that inscribed it.

With AI, we're not just automating transmission. We're automating what looks like thinking itself. And in doing so, we're discovering, or should be discovering, that much of what we called thinking was actually something else: pattern-matching, template-following, sophisticated mimicry. Real thinking, it turns out, is rarer, and more difficult than we wanted to admit.

This is both terrifying and liberating. Terrifying because it means many of the skills we've spent vast resources teaching are becoming obsolete before our eyes. Liberating because it forces us to ask what we should have been asking all along: what is education actually for?

The young people I work with will enter a labor market utterly unlike the one their parents navigated. Not because all jobs will vanish - I'm skeptical of such apocalyptic predictions - but because the relationship between credential and capability is fracturing. A degree proves you could navigate an institution for four years. It no longer proves you can think in ways a machine cannot.

What can a machine not do? It cannot care about the answer. It cannot stake something on being right. It cannot feel the particular frustration of being stuck on a problem, the breakthrough of sudden understanding, the deeper confusion that comes from learning enough to realize how little you understood before. 

It cannot want to understand differently tomorrow than it does today.

These human capacities - curiosity, intellectual courage, the ability to sit with uncertainty, the drive to understand for the sake of understanding - have always been what real education cultivates. But we've been so busy measuring proxies that we've barely noticed when the proxies became the point. An AI can write a brilliant practice essay about the themes in Toni Morrison’s Beloved, but only a human can be changed by reading it.

So, what should parents and students be thinking about as they navigate these college decisions in this peculiar moment?

First, perhaps, that the institutions most worth attending are those wrestling genuinely with these questions rather than pretending the ground isn't shifting beneath us. Where are the universities rethinking assessment, reimagining what it means to demonstrate learning, grappling with the difference between performing understanding and possessing it? These places exist, but they require looking past rankings and prestige markers that were designed for a different era.

Second, that the skills worth cultivating are increasingly orthogonal to what's easily measured. The ability to ask better questions. The capacity to synthesize across domains. The judgment to know when an elegant-sounding answer is wrong in subtle ways. The intellectual character to do difficult thinking when you could easily offload it. None of these appear on transcripts, yet all of them determine whether education will be truly transformative or merely credentialed.

Third, that the relationship between learning and employment is being rewritten in real-time, and no one, not college admissions offices, not career counselors, not the students themselves, knows exactly what the next chapter looks like. Which argues for depth over breadth, for developing genuine expertise in how to learn rather than accumulating credentials that prove you learned specific things that may not matter.

The Cambridge philosopher Michael Oakeshott wrote of education as "a conversation between generations," each adding their voice to an ongoing exploration of what it means to be human. But AI doesn't join that conversation. It only produces statistically likely responses based on the conversation it has observed. This distinction is everything. The question is whether our educational institutions can recognize and cultivate what is essentially, and irreducibly, human about learning.

I often think of my students who will graduate college around 2029 or 2030, entering a world where the baseline of what machines can do will be vastly beyond what we see today. What I want for them is not just to be employable, though I hope they will be, but to be intellectually sovereign. To know what they think and why. To recognize when their thinking is their own and when they're reproducing something they've absorbed. To understand what problems are worth solving, not just how to solve them.

This requires a different kind of education than most of us received. It requires institutions willing to value what's difficult to measure. It requires teachers who understand that their role is not to transmit information - machines do that brilliantly - but to provoke thinking, model intellectual courage, and help young minds discover what they're capable of when genuinely challenged.

Perhaps most fundamentally, it requires students who understand that the point of education is not to collect credentials but to become someone capable of thinking that matters in ways machines never will.

So, as your children decide where to invest the next four years, I'm curious: are they asking which institution will help them think in ways AI cannot? Are they looking for places that will challenge them to discover what's irreplaceable about their own minds? Or are they optimizing for brands and rankings that may already be measuring the wrong things?

It’s New Year’s Eve on the west coast of the US. We stand tonight at one threshold and willl wake up tomorrow at another, the turning of a year, yes, but also the turning of something larger, a transformation in what it means to be educated, to work, to think. I wonder what conversations you'll have as the year changes, what wisdom you're finding as you navigate this peculiar moment with your children. What resolutions might we make, not about weight loss or productivity, but about cultivating the kinds of minds that will flourish in a world we can sense but not yet fully see? These aren't rhetorical questions. I would love to know what you're discovering within your own realms, at this particular turning point, about what matters most.

Wishing you a fulfilling New Year!