The easiest post to write this fall would be the one against it. I could open with a warning about screens, line up every worry about AI, about cheating, about the slow erosion of thinking, then call for a ban, promise to protect childhood and close with something stirring about the good old days. It would be a popular post. It might be my most shared one all year.

I am not going to write that post.

Not because those worries are wrong. A lot of them are right. There is a real mood out there this fall, a pushing back against technology in schools, and much of it is healthy. We restricted phones. We are talking again about attention, about presence, about what it means to actually be in a room together. I am glad we are having those conversations. This blog has spent more than fifteen years arguing that the human part of school is the whole point.

The booing

Think back to this past spring. At college and university commencement ceremonies across the United States, graduates booed. Not at anything you would normally expect. They booed at AI. At the University of Central Florida, a speaker described AI as the next industrial revolution and was met with jeers. At the University of Arizona, former Google CEO Eric Schmidt was repeatedly booed as he discussed AI. At Glendale Community College, an AI system used to read the graduates’ names skipped some of them entirely, and the college president was booed when she explained what had happened, which honestly feels like the technology earning its own reception.

I keep thinking about that sound, and I do not think it was simple. Some of it was fear. This was a graduating class walking into a difficult job market, told over and over that a machine might take the work before they ever get to it. Some of it was the overhype, the sheer exhaustion of every product, advertisement and keynote speaker insisting that AI changes everything. Some of it was resentment at watching a handful of companies wedge themselves into every corner of our lives. Some of it was grief for something human that feels as though it is slipping away. And some of it, I suspect, was much simpler than all of that. It was young people, on one of the most human days of their lives, asking the adults at the podium for one afternoon, just one, where nobody mentioned the robots.

All of that was in the boo. That is what makes this hard. I do not think we can dismiss it as ignorance or resistance to change. It felt more like a very human response to being repeatedly told what the future is going to look like, often by people who have a significant stake in that future arriving.

And maybe that is part of what has me thinking differently about this year. I want school to feel more human, not less. More conversation, fewer screens. More looking up, less scrolling. More opportunities to make things, move, argue, wonder and be fully present with one another. If you have read me for a while, you know the word I chose for this year is Alive. That is what I want our schools to feel like.

It is also making me wonder whether we sometimes confuse being ready for the future with being ready for the newest technology. They are not the same thing. AI is certainly part of the world our students are entering, but surely the bigger goal is young people who are curious, adaptable and creative, who can think critically, work with others, make good judgments and learn something new when the world changes around them. AI happens to be one of the things forcing us to think about those qualities right now. It will almost certainly not be the last.

The world is not AI free

I understand the instinct behind a ban. Sometimes the easiest way to protect something is to keep something else out. But AI is not waiting politely at the school gate. It is already in the backpacks, on the phones and in the homework that gets finished at the kitchen table long after the bell. Our students use it. Their parents use it. I use it, and I tell you so at the bottom of every post. The idea that we can draw a clean line around the school day and declare it AI free has some appeal. But that line does not really exist anymore, if it ever did.

And this is not only an American story. A KPMG survey last fall found that 73 percent of Canadian students aged 18 and older were using generative AI for their schoolwork, up from 52 percent two years earlier. Whatever we decide about the place of AI in schools, students are already making decisions about it themselves.

That does not mean it belongs everywhere. In fact, the more I think about this, the more convinced I am that there are times when learning should simply be AI free. Some assignments, some conversations, some assessments and long stretches where students are expected to read, think, write, solve, struggle and create on their own. Young people still need to learn to write a sentence, develop an argument, sit with confusion and produce something that is genuinely their own. Sometimes the struggle is the learning, and if we remove the struggle we may also remove the thing we were trying to teach.

Context matters too. A six year old and a seventeen year old probably should not have the same rules. A student learning a skill and a student who has already developed that skill might reasonably have different access to tools. This is where I find simple answers less satisfying. There are times when restricting AI makes complete sense. There are other times when using it might deepen learning. Figuring out the difference seems much more useful, and much harder, than simply deciding whether AI is good or bad for schools.

That may also be one of the reasons school matters so much in this conversation. We spend a lot of time helping young people question sources, weigh arguments, recognize bias and tell the difference between knowing something and simply producing a good looking answer. AI complicates all of that, but it also gives us another context in which to teach it. A polished answer is not necessarily a thoughtful answer. Confidence is not accuracy. Generating something is not the same as understanding it.

A ban can establish a boundary. I am just not sure it can do all the teaching we need.

Even the builders are asking for caution

Here is another piece that keeps me from becoming much of an AI cheerleader. At the end of July, more than a thousand employees of frontier AI companies signed a statement called Pacing the Frontier. They did not call for AI development to stop. They asked the United States government to support an international effort to build the technical and governance tools that could deliberately pace frontier AI development if it begins moving faster than society can safely understand and manage.

That distinction is interesting. It is not a call for a ban, and it is not quite panic either. It is closer to asking whether we should make sure there is a brake pedal before discovering that we need one. Some of the people closest to these systems worry that the technology could begin advancing beyond our existing capacity for safety, oversight and governance. At the same time, many of them believe AI could contribute to a dramatically better future. Both ideas can apparently be true at once.

You could also fairly read a letter like this as its own kind of hype, insiders talking their book, and maybe some of it is. I do not know. But I find the uncertainty itself worth paying attention to. If some of the people building these tools are saying the outcomes are not guaranteed, it makes me a little wary when those of us in schools talk as though the path ahead is obvious.

Schools are certainly not responsible for solving the future of artificial intelligence. But we do have a role in helping young people live thoughtfully in a world shaped by powerful forces they did not choose and cannot entirely control. That feels like a bigger and more human challenge than teaching them how to write a good prompt.

The grey

Most of the interesting parts of life are grey, and this is no exception. A few years ago, I wrote that AI in education was getting murky. It has not gotten any clearer since. It has only gotten more interesting.

Perhaps the wrong question is whether technology belongs in learning at all. Schools have always been full of things that stand between a student and an idea: books, pencils, worksheets, screens and, sometimes, other people. Every tool mediates learning in some way. What I find myself wondering more now is which of those things deepen attention, understanding and relationships and which quietly diminish them. Which tools give students greater independence, and which make them more dependent? Which tools open possibilities, and which simply make it easier to avoid the work? Which tools help a teacher see a student more clearly, and which put one more layer between them?

I do not think those questions have simple answers. They depend on the student, the teacher, the task and what we are actually trying to accomplish. But I also wonder whether the more capable our technologies become, the more valuable some very old human capacities become. Curiosity, creativity, judgment, the patience to sit with uncertainty, the willingness to change your mind when the evidence changes, and the ability to make something with another person that neither of you could have made alone. We sometimes call these soft skills. There is nothing particularly soft about them. They may turn out to be some of the hardest and most important things we teach.

I have also seen enough by now to know this is not only a defensive conversation. I have watched a teacher use AI to clear away administrative work that was keeping her from her students, and spend that reclaimed time actually with them. I have watched it give a student a second way into an idea the first explanation did not unlock. I have watched young people use it to trace a claim back to its source, and push a draft they had already started further than they would have managed alone. None of those examples makes me think we should use AI everywhere. They do make it harder for me to argue that it has nothing useful to offer.

The version of this I keep circling back to is one I have written about before. What if AI, of all things, could make school less about screens and more about people? What if handing off some of the genuinely tedious digital work, without handing off the thinking, gave teachers and students more time to talk, connect, build, move and look one another in the eye? I believed that was possible when I wrote it. I still do.

Of course, that does not happen automatically. Technology has a way of filling whatever space we give it. Getting to a more human version of school will require some intention, some professional judgment and probably some mistakes along the way. It will also require us to be willing to say yes in some places and no in others, and to change our minds when the evidence tells us we should.

Maybe a better question

A couple of years ago, almost every conversation I heard about AI in schools was about what it can do. The demonstrations were irresistible. Look what it can write. Look what it can draw. Look what it can create in ten seconds. There was an understandable fascination with the party tricks.

Lately, I have noticed the conversation changing. Not everywhere, and not all at once, but increasingly I hear people asking not only what AI can do, but what it should do. And, perhaps even more importantly, what it should not do.

I like that shift. It feels less like adoption and more like education.

Maybe that is ultimately what I mean when I talk about preparing young people for the future. It cannot mean accurately predicting what the future will look like, because history suggests we are not particularly good at that. It might instead mean helping young people become comfortable making choices in a world that keeps changing. Choices about which tools to use, which claims to trust, when to ask for help, when to struggle on their own and when to reject something even though everyone around them seems to be embracing it.

AI makes those questions urgent, but they are not really AI questions. They are questions about judgment, independence and agency.

Where I am landing, for now

So perhaps that is where I am at as we begin another school year. I do not want us to pretend AI does not exist, but I also do not want fascination with the technology to become the point. I want us to stay curious about where it might help, cautious about where it might get in the way, and protective of the parts of school that remain deeply human: conversation, mentorship, relationships, belonging, connecting and the experience of working through something difficult together.

None of this feels particularly settled to me, which is probably appropriate. I am less interested than I was a couple of years ago in predictions about how AI will transform education. I am more interested in the smaller and probably harder questions about what we value, what we want students to be able to do for themselves, and where technology helps or undermines those goals.

So I will end where I usually do, with questions. What does a more human school actually look like this year? Where does AI help us get there, and where does it quietly get in the way? When should students use it, and when is the most important lesson learning to continue without it? What capacities will matter most for young people in a future none of us can predict very well? And how do we make sure the time technology saves gets spent on one another, and is not simply swallowed by the next screen?

I do not have all the answers. I am becoming increasingly doubtful that anyone does.

But school still seems like a pretty good place to work through the questions.  

Welcome back.

Note: The image at the top of this post was generated using AI. I also used AI tools for editorial feedback while refining the piece. The arguments and conclusions are my own.