How to build the classroom of the future

NOTE FOR 2021 READERS: This is the 18th in a series of award-winning open letters to the next century, now just one generation away. Kids born this year in the U.S., and nearly 50 other countries, are expected to live to 2100 and beyond. Dear 22nd Century examines likely scenarios for their future, and how we can make the best one happen.

Dear 22nd century: So, how was school today?

Of all the areas where we could measure progress between the 2020s and the 2100s, none seems more up in the air right now than education. The pandemic forced us into history’s largest experiment in remote learning, from kindergarten through college — and beyond, as Massive Open Online Courses (MOOCs) like Udacity, Coursera and EdX exploded in popularity during the early months of lockdown. COVID struck a blow to standardized testing, with an incredible two thirds of colleges dropping their SAT and ACT requirements. You don’t know what those acronyms mean? Good. Let their systemically racist, eugenicist legacy remain in the dustbin of history.

If you were a starry-eyed optimist with a desire to disrupt our lumbering old system and an ed-tech hobby horse to ride, the return to school in fall of 2021 was your moment. Funding for education technology startups around the world has nearly tripled in two years. We seemed on the cusp of a hybrid education world — high tech and low tech, remote and in-person, in a ratio tailored to each student’s needs. Perhaps in the near future a thousand classrooms would bloom, and the marketplace of ideas would spread the most effective new models around the world. Childlike curiosity would be unsuppressed, the joy of learning would be universal, and generations of sub-optimal outcomes would end in the blink of an eye.


“Educators built an exact digital replica of what they were doing before.”

On the other hand, the remote learning experiment of 2020-21 was somewhere between disappointing and disastrous for many families — those at the bottom of the economic pyramid especially. Students in low-income schools were seven months behind in math last year, a McKinsey study found, two months worse than the average for all kids. Meanwhile, judging by the parents on my Facebook feed, plenty of teens are taking the SATs even when they don’t have to — because extra credit.

“We all know SATs aren’t really optional to get into a top 40 school,” one mother tells me. On the advice of guidance counsellors, she’d just spent $4,000 on tutors that would give her son the best shot at a $100 SAT test.

“What was crazy about the pandemic is that through blood, sweat and tears, educators kind of built an exact digital replica of what they were doing before,” says Justin Reich, director of the MIT teaching systems lab and author of Failure to Disrupt: Why Technology Alone Can’t Transform Education. “Very few places saw structural changes to how they organize schooling.”

The folder in the backpack may have been replaced by a folder in Google Classroom, the textbook may be on your iPad, but class still lasts for about 50 minutes and the teacher still want to see you in your seat. The basic Victorian model of education — bored kids squirming at desks, teachers and textbooks invested with god-like authority — is alive and well and living on the internet.

View it in this light and the chances of the next few generations of educators smashing this system, instead of just tinkering with it or providing alternatives for high-income families, seem slim indeed.

Maria Montessori, education revolutionary, in 1946. Her ideas are still not widely accepted.

Maria Montessori, education revolutionary, in 1946. Her ideas are still not widely accepted.
Credit: Kurt Hutton / Picture Post/Hulton Archive / Getty Images

It isn’t just the pandemic, you see. We’ve been trying to disrupt the Victorian model with limited success for more than a century now. Maria Montessori opened her first enlightened classroom in 1907; the system she started encouraged play and creativity in learning and discouraged grades. Sticking a letter from A through F on homework does not help foster learning in the long term, as the Montessorians have insisted for decades.

And yet, even with studies showing the effectiveness of their methods, there are still only 5,000 Montessori programs in the U.S., and only 500 are within public schools (of which, to give you a sense of scale, there are 131,000).

Every generation seems to offer its own new model that is going to upend education: Summerhill school in the UK, founded in 1921 as a boarding school governed by its pupils; Sudbury Valley school, another democratic institution founded in Massachusetts in 1968; the “unschooling” movement (think homeschooling meets Montessori) from the 1970s on. Each survived. Each has its adherents and imitators. None do more than nibble around the edges of the vast public school system. They’re the exceptions that prove the rule.

Nor is our obsession with education technology as new as we think. “Books will soon be obsolete in the schools,” famed inventor Thomas Edison declared in 1913; motion pictures would replace them “inside of 10 years.” When his deadline was up in 1923, speaking to the FTC, Edison upped that estimate to 20 years hence. If Edison had written a “Dear 21st century” letter, it would be our sad duty to reply that a class consisting entirely of video is widely seen as proof that the teacher has checked out.

Reich has a telling story about one high school principal who bought piles of physical textbooks, even though his classes were trying to move to digital versions with online videos and quizzes, because most of his kids’ parents expect to see books when they visit. It was easier to build a Potemkin village of expensive textbooks than proudly tout new tech. That sound you hear is Edison spinning in his grave.

Sebastian Thrun, self-driving car pioneer, Udacity co-founder, and perpetual Silicon Valley optimist.
Credit: Michael Macor / San Francisco Chronicle via Getty Images

We don’t seem to learn to temper expectations. A century after Edison’s first textbook prediction, in the early 2010s, we were in the grip of Silicon Valley mania about how the rise of the internet would transform the entire educational landscape. When former Google VP Sebastian Thrun co-founded Udacity, he tells me, he thought MOOCs might cause massive consolidation, to the point where 10 companies or institutions would provide half of all higher education within half a century.

“I would probably have to recognize that the willingness of existing universities to leverage technology has been less than I hoped,” Thrun sheepisly admits. The company has shifted focus to the international scene — we’ll get to that — and to corporate training courses.

Udacity’s one bright shining exception to the rule was Georgia Tech; its online computer science master’s degree, developed with Udacity, was the top ranked computer science degree in the U.S., while enrolling an incredible 10,000 students in 2020. But that was the year Georgia Tech took the service off Udacity and onto its own cheaper platform. The more popular university-led MOOCs are, the more universities are going to want to control them. (You can still access some of the classes on Udacity.)

It’s not that the disruptive ideas are flawed in themselves. It’s that even the best big dreams run into the rocks of an extremely conservative system on their way to changing the world. Why conservative, if we all loathe the current state of schooling (Gallup polling over the last 20 years shows that roughly half of the U.S. has been “somewhat” or “completely” dissatisfied with K-12 education for all that time), if so many of us have better ideas for how to run a classroom?

Well, says Reich, it’s probably because our better ideas don’t align with each other. We’d rather our kids went through the flawed system that we ourselves survived than submit them to something untested and unknown.

“You know, as a researcher, I sort of like the phrase ‘experiment on children,’” jokes Reich. “But it turns out that most people don’t.”

Cloud school, from both sides now

Sugata Mitra, standing at right, visits his “School in the Cloud” in New Delhi in 2014. On screen, a member of the “granny net” is the only adult supervision.
Credit: Ramesh Pathania / Mint via Getty Images

When I think of ed-tech pioneers with big appealing dreams that got downsized, I think of Sugata Mitra. In 1999, Mitra was the chief scientist at the National Institute of Information Technology in New Delhi. The school was next door to a slum, and Mitra decided to conduct what became known as the “hole in the wall” experiment, where an internet-connected computer was made available for kids in the slum who’d never seen one before — literally stuck in the wall, at a child-friendly height.

To Mitra’s delight, kids in this and later experiments not only taught themselves computer literacy by trial and error, but also figured out how to use email, instant messages, and, crucially, search engines. He tried loading one hole-in-the-wall computer with educational materials on molecular biology; sheer curiosity, along with encouragement from an adult supervisor, was enough to give them test scores that put them on a par with kids in an elite New Delhi school.

Mitra’s experiment was the inspiration for the novel Q&A, which became the Oscar-winning movie Slumdog Millionaire, which turned on widespread incredulity that someone born so low could possibly figure out how to garner knowledge from the internet.

That success, since replicated in more than 300 “hole in the wall” experiments across India and sub-Saharan Africa, made Mitra an ed-tech star. I met him at the TED conference in 2013, where he was awarded $1 million for his next vision: the “school in the cloud.” These were minimalist classrooms where kids clustered around a single computer and were posed one deceptively simple question at a time, to be answered collaboratively.

As you can see in a 2018 documentary, The School in the Cloud, questions had a whimsical but genuinely big-picture nature; a genre that should be familiar to, well, anyone who’s heard children ask questions. Why is the sky blue? Why is ice slippery? Can trees think? Does tomorrow exist? (You are proof that the answer is yes, dear 22nd century, so don’t let me down by not existing.)

No teacher would be present, but a team of volunteer retirees from around the world — Mitra called it the “granny cloud” — would provide a gentle mix of encouragement and delight at the kids’ discoveries via video. In the optimistic early 2010s, it seemed an intriguing and plausible experiment.

I hadn’t seen any news about what happened next, so I caught up with Mitra, now retired himself and, randomly, living near the town where I grew up in the North East of England. Turns out he spent his prize money on seven schools, five in remote areas of India and two in the UK. They were beautiful glass hexagonal structures, complete with solar power that would keep them running for almost nothing. But the communities they were in would still be responsible for the upkeep. Would they see enough value in it?

You can probably guess the answer already. Despite tentative evidence that cloud-school kids “tend to answer more challenging questions and retain the information for a longer time than they would usually,” in the words of one academic review of Mitra’s work, it wasn’t designed to make them pass tests. “What the communities in India said was, ‘this is not going to do anything for the school certificate examinations,’” Mitra says.

The cloud schools in England fared better, but parents decided to stop sending their kids around 14 — when British kids have to start preparing for their GCSEs (exams taken by UK kids at 16).


“We Google all the time. Why don’t we let the children Google?”

When Mitra pointed out that reading comprehension should help anyone prepare for any exam, the parents shot back: yes, but in an exam, they’re supposed to answer alone. “And that word, ‘alone,’ was the showstopper,” he says. “Because my method was the opposite of alone.”

So Mitra scraped together money from speaking gigs and tried again. This time, he built cloud schools as part of a broader curriculum inside other schools; one in Harlem, one in a village in West Bengal.

This time, actual paid, trained teachers played the role of the granny cloud. The addition of in-person supervision made all the difference. “I didn’t need to explain very much, the teachers got the hang of it,” says Mitra. “What they realized was, ‘well, we Google all the time. Why don’t we let the children Google?’”

It’s a fair point. Why do we persist in pretending the internet isn’t a thing? When are future…

Leave a Comment