Neil Postman, the media critic, spent the 1980s warning people that America was raising generations who believed themselves powerless before new technology — not because the modern advances were overwhelming, but because the adults kept responding to it in an uncoordinated scramble.
Dartmouth engineering professor Rafe Steinhauer told the Valley News that he thinks about Postman’s arguments frequently as he grapples with today’s technology.
Steinhauer, who studies how schools make decisions, thinks that whatever Upper Valley schools do regarding artificial intelligence and screen time, they have to do it with conversation and a plan. By organizing with schools, parents, students, and municipalities, Steinhauer hopes to help both New Hampshire and Vermont make informed decisions as a community.
“There’s an implicit opportunity to teach our students that democratic processes still have power over emerging technologies,” he said, “at a moment when our national leaders are shirking that responsibility.”
Two schools on either side of the Connecticut River are taking that opportunity — one with Dartmouth’s help, one without — to work on the problem from different ends.
At The Sharon Academy in Vermont, students’ phones have been placed in a locked cabinet each morning for the past two years. This fall, personal laptops will join them — barred entirely and replaced by five carts of school Chromebooks a teacher wheels out only when lessons demand.
Across the border in the Lebanon School District, which is working with Steinhauer’s team, educators decided to leave general screen time alone and spend the next two years writing careful rules for one of the most powerful things those screens can do: run AI.
Both are answering the same question — what belongs in front of a child, and when — and Steinhauer said the two halves were never really separable.
“You can’t really have the question about AI without thinking about what the hardware and software they’re engaged with are already,” he said.
Lebanon schools’ Chromebooks run Google’s Gemini. Even with daytime cell phone bans, most teenagers carry a chatbot in their pocket at some point during the day, and plenty use one at home to get through assignments, much to the chagrin of their teachers. How much screen time a kid should get, and what that screen should be allowed to do, are two of the most prevailing questions in American education.
As of June, 39 states and the District of Columbia require districts to restrict student cellphone use; Vermont and New Hampshire are among those with bell-to-bell bans. New Hampshire’s Department of Education has since convened a School Screen Time Task Force to take up everything else with a screen — to determine whether the technology schools spent two decades adopting is helping or hindering.
Nobody knows yet.
Schools have gotten this wrong before, and fixed it
Sharon’s phone ban predates Vermont’s statewide mandate, which starts this school year, and Head of School Mary Newman said the effect was immediate.
“Quickly, that was a huge improvement, pretty much on all levels — academically, socially,” she said. “Students hanging out with each other and having conversations, and reaching out to teachers much more than they had in the previous years.”
Steinhauer said that when he was earning his master’s in education, the open question was what role smartphones should play in classrooms, and he wasn’t sure they should be banned at all. “Fast-forward many years, I’m a huge proponent of bell-to-bell bans,” he said. “The sophistication with which these technologies are designed to hack our attention is too strong.”
Because he was wrong about phones, he said, is why he has approached AI “with some humility.”
“The people who say we should just be banning it across the board — I don’t think they’re right,” he said. “But I’m not sure they’re wrong, either. And the people who say our students will get left behind if they don’t learn these tools — I don’t think that’s exactly right. But I can’t say confidently it’s not.”
Either way, he’s convinced that the conversation has to happen out loud, and it’s getting harder to have.
“It’s getting pulled through our left-right coding in American politics really quickly,” Steinhauer said. A couple of school leaders have told him the aversion among some teachers to even discussing AI has grown — their speculation, he said, being that “AI now codes right [wing] in a way.”
This fall, Steinhauer and Adam Norwood of the Upper Valley Educators Institute are inviting school and district leaders from both states to start meeting about AI, through Dartmouth and UVEI.
Steinhauer said one teacher replied to the invitation with an answer he described as flatly hostile. Steinhauer said he isn’t promoting AI, but a discussion about it.
“There’s a lot of fear. There’s a lot of concern. There’s a lot of skepticism,” Steinhauer said. “A lot of it is merited.”
The same question, harder
Humility matters more this time, because Steinhauer doesn’t think AI is just the next smartphone.
“AI is the future of software,” he said. “That doesn’t mean it’s the future of education. And it shouldn’t be, in my opinion.”
He measures it against personal computers, the internet, the smartphone — each its own reckoning, each survivable. But one thing here has no precedent: “None of them, with a very lightweight prompting, could generate an answer in the voice of the student, instantaneously, and almost at any length.”
A machine that writes like a specific child, on demand, is not the same problem as a machine that distracts her.
His own worry about AI isn’t cheating – it’s what cognitive scientists call verbal formulation — the ordinary work of turning a thought into language.
“Our ability to put an emotion or thought into language not only underpins our ability to have dialogue in society, but also a lot of aspects of mental health, relationships, and other things,” he said. He draws a straight line between smartphones and the collapse in attention spans that nobody flagged in time.
Nobody agrees, including within buildings
At Sharon, the trouble surfaced as soon as the phones were gone and laptops filled the space they left.
“Students were turning to their laptops in a really similar way that they had been turning to their phones,” Newman said.
Dean of Students Blake Fabrikant started paying attention after reading a book called “The Digital Delusion,” which cites a figure he said he now repeats often: for every hour a student is supposed to be working on a screen, 38 minutes are off task.
“I would see students playing video games when they were supposed to be researching,” he said.
So this fall, when the laptops go away, “you feel socially anxious, you can’t turn to a screen,” Fabrikant said. “You have to take a walk. You can talk to a friend.”
Sharon has no written AI policy yet, which hasn’t stopped Fabrikant and Newman from arguing about it.
Fabrikant’s said a year ago he thought: “Get it out of here. It’s as bad as vaping.”
He’s since come around to supervised use — a statistics project, say — but not to students working with it on their own. “They need to learn how to think, not what to think,” he said.
Newman would rather teach about it than ban it.
“It’s way more nuanced than just a flat no,” she said, “because AI is way bigger than something you can just say yes or no to.”
Lebanon came at the same worry from the other side: leave screen time alone for now while they write the AI policy.
“We knew right away that generative AI was going to be really transformative to the landscape that we work in,” said Karen DuBois, the district’s director of academics, engagement and opportunity.
Students test-drive Gemini “in very specific use cases,” she said, in grades seven through 12. Below that, it’s switched off; when it’s on, it’s “always used for specific guidance under the direction of a teacher, with their permission.”
Ask DuBois whether AI is simply the next thing after laptops and smartboards, and she hesitates. “Characterizing it as simply a tool like any other tool may be an oversimplification,” she said. What worries her is cognitive offloading — “when we are shortchanging our own selves through the use of AI tools.”
The split runs through individual buildings, not just between districts. Some teachers are building AI assistants into daily lessons, with bots in the classroom to help students better understand what they know and don’t. Others intentionally do their best to avoid AI and plan to keep it that way.
Steinhauer said to make the best move for everyone involved, “schools need to coordinate for the sake of having the best decisions about AI.”
The people who haven’t been asked
This fall, Steinhauer and Norwood are inviting district leaders from both states into a working group with Dartmouth, so schools can decide these things together rather than each guessing alone. It’s built on a premise he teaches constantly: that power in a school is diffuse, that any group left out — teachers, parents, students — can quietly sink a decision made without them.
The group he’d most like to hear from is the one that rarely gets asked.
“I’d like to see more student leadership on this,” he said. Students “not only can learn that they can play an important part in school decision-making, but I think their values matter a lot.”
Newman has noticed those values arriving fully formed. A growing number of her students, she said, “are coming into high school already really anti-AI.”
Their objection isn’t the one the adults keep having. “They care about the environment,” Fabrikant said. “And then also just the idea that you might not have jobs — that is really terrifying.”
The kids know AI is here. Steinhauer wants the adults to prove they can handle it.
“Schools are the most important institutions, and children are our most important people,” he said.
“That’s the paradox of the work: how do you build a culture that is like, yes, this really matters what we do, and therefore we’re going to be willing to experiment and learn as a collective even more.”
Wait for certainty, and it arrives long after this generation has graduated.