Matt Barnum is interim national editor of Chalkbeat. This article appeared October 4, 2023 in Chalkbeat.

Several years ago, Mark Zuckerberg had grand designs for American schools.

The Facebook founder and his wife, pediatrician Priscilla Chan, poured well over $100 million into an online platform known as Summit Learning that initially aspired to be in half of the nation’s schools. In 2017, Zuckerberg suggested that technology-based “personalized learning” could vault the average student to the 98th percentile of performance.

Fast forward to this summer: The Chan Zuckerberg Initiative, the couple’s philanthropic arm, laid off dozens of staff on its education team and announced a shift in strategy. “Our understanding of what’s possible in the world of education — and in our world more generally — has changed,” Sandra Liu Huang, CZI’s head of education, wrote in an August blog post. “And so, at CZI, our education efforts must change too. Navigating these changes is humbling and challenging, but ultimately, necessary.”

It was an acknowledgement that the company’s prior education strategy had fallen short of its hopes. Through a spokesperson, Huang declined an interview request, but noted in her blog post that the company is continuing its work in education, albeit with a different strategy. “This moment demands not just investment but innovation — and that’s why we are building a team of experts and partners to identify opportunities where technology and grantmaking can drive coherence,” she wrote.

CZI’s shift in approach marks something of a coda to an era when various advocates and funders believed that computer-based “personalized learning” could dramatically improve education. Summit, CZI’s pet project, has not spread as far as once hoped, and there’s little evidence that it or similar efforts have led to the large learning gains that Zuckerberg envisioned. This gap between ambitions and results underscores the difficulty of using technology to dramatically improve America’s vast system of decentralized schools.

“People keep hoping that our technologies are the Swiss Army knives or steamrollers that they can do everything,” said Justin Reich, a professor at MIT and author of a book on the limits of technology in education. “Instead, our best technologies are very particularly shaped ratchet heads and the landscape of education is millions of bolts.”

A big bet on “personalized learning”

In 2014, Zuckerberg and Chan both visited Summit Public Schools, a Bay Area charter school network that had recently integrated technology into the school day. Zuckerberg was impressed and wanted to know how to scale their model nationally.

The following year, Facebook announced a partnership with Summit to help develop software that would be made available outside the charter network. “It came onto the scene with quite a bit of fanfare,” said Sarah Reckhow, a professor at Michigan State University who has studied education philanthropy. 2015 was also the year that Zuckerberg and Chan pledged to donate 99% of their Facebook fortune to charity, through the newly formed CZI.

Summit would be a key part of the couple’s education strategy. A case study from the time described Summit as a multipart model, which included time that students spent with an adult mentor, project-based learning, and “expeditions” in various fields like arts or computer science.

Summit also featured 16 hours a week of “personalized learning time.” Students worked at their own pace on a computer, which fed them a “playlist” of content where they learned specific skills. Students could move on once they got eight of 10 questions right on an online quiz.

That seemed to be the biggest draw for Zuckerberg, who contrasted the approach to “having every student sit in a classroom and listen to a teacher explain the same material at the same pace in the same way.” He suggested this could lead to transformational improvements in student learning. The goal, he wrote in 2017, was “scaling this approach to every classroom.”

Zuckerberg was also a part of a $100 million investment in AltSchool, a private network of schools run by a former Google employee who hoped to spread innovations to the public schools.

At that point, the excitement was based more on hope than hard evidence. Still, Zuckerberg and Chan were hardly the only enthusiasts for technology-infused learning. Other funders, including the Gates Foundation, promoted similar ideas. (Gates is a supporter of Chalkbeat.)

“Personalized learning” had emerged as “a reliable buzzword in education circles, and right now, its star is rising,” wrote Michael Horn for Education Next in 2017. “Leading foundations, school networks, and nonprofits are putting it front and center in their efforts.”

Summit comes under scrutiny

Over the years, Zuckerberg ended up pouring large sums of money into Summit Learning. In April of 2019, The New York Times reported that CZI had made $99 million in grants to Summit. More recently, between 2019 and 2022 according to CZI’s grants database, the organization channeled $125 million to Gradient Learning, a nonprofit that now administers Summit.

These are vast sums in the world of education philanthropy.

Much of the money has gone to develop and maintain the Summit Learning Program and provide training to educators. Schools can receive the program and training free of charge.

Hundreds of schools have adopted Summit, and some parents say it’s helped their child. “Students who may understand a concept more quickly are able to move forward instead of having to wait for the rest of the class,” wrote Annie Thomas, a Colorado parent who defended the program in an op-ed last year.

But in other cases, parents complained about the newfangled approach to education or worried that students’ data would be shared with Facebook. (Summit says Facebook does not have any access to its data.) Some students said there was too much time on screens. In 2018, students at a Brooklyn high school walked out in protest, prompting a flurry of news stories. The following year, the New York Times published a piece featuring complaints from parents in Kansas. “We’re allowing the computers to teach and the kids all looked like zombies,” one parent told the Times.

It wasn’t clear how representative either the success or the horror stories were, though.

That’s in part because Summit struggled to produce evidence that its program was leading to the large learning improvement that Zuckerberg had hoped for. Summit declined to participate in a study of the program designed by Harvard researchers. (Asked about Summit’s effectiveness now, both CZI and Gradient Learning pointed to positive experiences in a number of schools, as well as a survey of school leaders in which nearly all said Summit had made a positive difference for their schools.)

Reich says the apparently varied experience of schools using Summit shows how contextual factors, like strong leadership or community buy-in, can make or break a school regardless of the technology. “Without all of the other things that the Summit Public Schools were doing to make that platform work, they were not scaling or transporting the most important parts of Summit,” he said.

Data previously obtained by Chalkbeat showed that 1 in 4 schools that began the program had dropped it by the 2018-19 school year, at which point just short of 400 schools were using Summit.

Presently, Gradient Learning says that over 300 schools are using Summit, indicating that in recent years more schools have left the program than have newly adopted it.

A post-pandemic pivot for CZI

In 2020, the pandemic thrust students online all across the country. It was a real-world experiment in the potential — and limits — of technology-based learning.

Zoom school was a struggle for many students and teachers alike. Test scores cratered. But the shift likely had the enduring effect of infusing more technology into the typical school day even when students returned in person.

More schools invested in laptops or tablets and tout that they are 1:1, meaning every child has a device. Some places have bought online learning programs in a bid to help catch students up. Some school districts have maintained fully online schools.

“Aspects of the technology-mediated ‘personalized learning’ do seem to be present in a lot of places,” said Reckhow. “It so clearly got accelerated by COVID.”

The scope and efficacy of these changes is not yet clear. But there’s little indication that this shift has delivered the large gains in student learning that Zuckerberg once touted.

And some of CZI’s own big bets have fallen short of its ambitions. By 2021, AltSchool had sold its schools and turned into a consulting firm. Most significantly, CZI now appears to be reducing its focus on Summit, apparently backing off of hopes that the model would spread to vast swaths of American schools. There are plans in place to “transition” the “core features of Summit Learning to a third-party platform,” CZI says.

What this means for Summit isn’t entirely clear. Both CZI and Gradient Learning have indicated that they are not giving up on the model altogether.

“We continue to have a firm commitment from the CZI team about their continued support of the Summit Learning program, in its current and future iterations,” according to an email sent to supporters by Gradient Learning, obtained by Chalkbeat.

Asked about CZI’s broader experience in education over the last several years, a spokesperson said, “Thanks to the work of our grantees and partners, more schools are embracing a whole-child approach to learning; educators increasingly have access to professional learning opportunities that incorporate the latest learning science research; and researchers and educators are developing new ways to measure student progress that more accurately reflect their full potential.”

CZI is not getting out of education. Huang, the company’s education head, said the company will now focus on “creat[ing] educational tools” that can help “address chronic learning and teaching challenges.” That includes exploring how artificial intelligence can improve education, according to a document obtained by Insider. (A spokesperson for CZI said in an email, “The public discussion about AI in education is evolving rapidly, and we are approaching it thoughtfully and cautiously.”)

John Bailey, a fellow at CZI and the American Enterprise Institute, recently wrote an optimistic essay about the potential of AI, with a headline that marked the end of one era and the dawn of a new one: “The Promise of Personalized Learning Never Delivered. Today’s AI Is Different.”