It started with an email from Yong Zhao. March 2020. Schools were closing. The question was simple and enormous: What if schools were closed for a year — what would happen to learning? A group of us got on a call to talk it through — Yong Zhao, Chris Dede, Curt Bonk, Punya Mishra, Scott McLeod, and Shuangye Chen — and by the end of the conversation, we'd decided to do something about it: a weekly webinar, open to anyone, starting immediately. Chris gave the series its name: Silver Lining for Learning. The first episode aired on March 21, 2020, the very week the world shut down.
As life changed, so did the team. Scott and Shuangye eventually stepped away, and Lydia Cao joined, bringing new energy and research depth. The current hosting team — Chris Dede, Curt Bonk, Lydia Cao, Punya Mishra, and Yong Zhao — has been in place since.
The pandemic itself was only the beginning. In the years that followed, schools confronted cellphone bans and the collapse of attention, governments moved to regulate social media for young people, generative AI went from a research curiosity to a force reshaping every classroom conversation, and political turbulence upended education systems worldwide. Six years is not a long time. But these six years compressed several generations of change into one.
The name turned out to be apt in ways no one intended. Chris's original optimism — that the pandemic might be the push online learning had always needed — proved more complicated than anyone expected. But the series itself became the silver lining: a sustained space for thinking together when everything else was in upheaval, and a platform for surfacing the people and organizations doing extraordinary work in education that too often goes unnoticed. That combination is what kept the show alive long after the emergency that launched it had passed.
SLL is entirely unfunded. No grants, no sponsors, no institutional support. Five busy academics — scattered across universities, time zones, and continents — find the guests, prepare the sessions, host the conversations, edit the videos, post the podcasts, and maintain the website. Every week, for six years. It is, in the most literal sense, a labor of love.
Six years and 265 episodes later, the series has featured guests from over 30 countries, spanning K–12 classrooms, universities, ministries of education, NGOs, and the technology frontier. Along the way, it has been recognized with AECT's Distinguished Development Award (2022), an Outstanding Digital Learning Artifact award (2022), and an Annual Achievement Award (2023). But the better measure of SLL is the range of voices it has brought together: students and scholars, designers and policymakers, classroom teachers and technology pioneers, all sharing their expertise and vision for what education could become.
SLL would be nothing without the generosity of its guests — hundreds of people who gave their time freely, often joining from across the globe at improbable hours of the day and night, because they believed this conversation mattered.
The Accidental Dataset
SLL was never designed as a research project. But 264 episodes of recorded dialogue between hosts and guests, fully transcribed, amount to over 2.6 million words of continuous discourse about education during one of its most turbulent periods. (One recording, episode 166, was lost to a technical failure, so 264 transcripts survive from 265 broadcasts.) That's not just a series. It's a dataset.
This site is our attempt to look at that dataset systematically. A close analysis of the full corpus revealed twelve recurring themes, organized into four clusters that reflect the fundamental arc of the SLL conversation. But the themes are only the beginning. The deeper you look, the more the data reveals: structural shifts over time, unexpected alliances between ideas, moments where the conversation changed in ways that no single episode made visible. Some of these patterns confirm what you'd expect. Others are surprising.
The analysis grew out of a process of dialogic coding — an iterative collaboration between the researcher and an AI interlocutor, described in detail on the Methods page. The next page explains how twelve themes emerged from the data, and what their structure reveals.
What This Site Offers
The pages that follow present the findings in four parts. The Story introduces the twelve themes and four clusters, and shows how hosts and guests emphasize different things. The Analysis offers seven interactive explorations — from a PCA landscape of all 264 episodes, through streamgraphs and keyword evolution, to Shannon entropy, correlation networks, and novelty scoring — each revealing a different structural feature of the conversation. The Synthesis steps back to ask what happens when all six analytical methods agree, and documents the methodology behind them. And Go Beyond opens the full data archive for download and introduces the people behind the show.
Each page can stand alone, but they're designed to be read in sequence — each building on the one before it, like chapters in a book about a conversation that hasn't ended yet.