すべてのエピソード

I tried to understand the significance of a shadow in the Russian Arctic. A few weeks ago, a strange-looking shadow on Google Earth sent me down an unexpected rabbit hole. I set out to figure out what it was — and discovered a fascinating story happening in the background of the race for the Arctic.

Inside a fake science factory. Whenever I want to understand the latest discoveries in a topic I’m reporting on, I turn to the same place: research papers. So when I first learned about “paper mills” — shadowy businesses that produce and sell fraudulent research to academics trying to meet publishing quotas — I was fascinated. Over the past few weeks I dove into the research, interviewed scientists, and spoke to a real paper mill representative to answer a troubling question: Can we actually trust the research papers we find online?

I reverse-engineered a Tinder scam. Some Tinder users have started noticing a strange new kind of profile. Swipe through the photos and everything looks normal — until you hit the last one. That's where things get weird. The final photo always shows a face that doesn't quite fit: sometimes silly, sometimes unsettling. I went down a rabbit hole of stolen identities and catfishing schemes to figure out what's really going on.

I tried to understand the Ray-Ban Meta glasses backlash. Over the past few years, Meta’s smart glasses have gone from a niche gadget to the world’s first mainstream wearable AI device. Millions of users can now capture moments, listen to music, interact with an AI assistant totally hands-free. But the same design that makes them so seamless — glasses that look like regular glasses — has also made them a tool for a new kind of harassment. Content creators have used them to secretly film strangers; women have found footage of themselves online with no idea they were ever recorded. I bought a pair to understand both sides: the genuine appeal of a device that might change how we interact with the world, and the very real harm it's already enabling.

Why the 0.1% dominate Kalshi and Polymarket. The internet can't stop talking about prediction markets. Kalshi and Polymarket — sites where you can bet on almost anything — claim to offer two things: an accurate window into the future, and a chance for anyone to make real money. Some people are winning big. But who are they? And what do they know that you don't? To find out, I dove into the latest academic research on how these platforms actually work. The story of who's really winning on prediction markets is more complicated, and more surprising, than Kalshi or Polymarket would have you believe.

We tried to figure out why it’s so hard to delete a David’s Bridal account. When my friend Kim tried to delete her David’s Bridal account, she ran into a strange problem. Online, the dress retailer didn’t have a way to delete an account. Phone calls with customer service were a dead end too. She reached out to see if I could figure out what might be going on. What followed was a months-long runaround that took me behind-the-scenes of personal information retention, data brokers, and exactly how much a dress company can know about you.

The myth hiding a real problem. It has, almost certainly, happened to you. You have a conversation in person, you mention something specific, and minutes later, there’s an ad for it on your phone. For the past decade, most of us have come to the same conclusion: Our phones must be listening. But are they, really? I wanted to dig into the “phone listening” theory — where it came from, what the official research says, and what’s actually happening every time you load an ad on the internet. What I learned is that creepy ads are just the tip of the iceberg, powered by a system auctioning off our location, identity, and behaviors to countless companies vying for our attention… and others looking for something else.

Who's actually behind the angriest content on Facebook? Facebook users across the US, UK, Australia, and Canada are discovering something strange on their feeds. At first glance, it looks like national news content: street interviews about hot-button issues, videos of protests, graphics with attention-grabbing headlines. If you know what to look for, you can tell that basically all of the content is AI-generated. But these news accounts are hiding an even bigger secret: they aren’t actually based where they say they are. AI slop accounts covering American, British, Australian, and Canadian news are actually being run by people in Sri Lanka, Vietnam, and Cambodia — and much of the content they post is specifically designed to make viewers as angry as possible. I dove into this bizarre corner of Facebook — speaking to misinformation researchers, AI-generated content creators, and a former Facebook data scientist — to understand why this flood of rage bait is taking over Facebook.

I teamed up with Bellingcat to explore why a 1996 murder fell through the cracks. 28 years ago in the quiet Boston suburb of Sharon, a dog came home carrying a human bone. Police followed the trail into a nearby forest and found a decomposed skeleton: bundled in quilts, 98 feet from the road, with two bullet holes in the back of the skull. Detectives worked the case for years. But if you look it up today on NAMUS — the national database of missing, unidentified, and unclaimed persons — he’s still listed as unidentified. When Bellingcat investigator Peter Barth heard about this case, he couldn’t let it go. Who was this person? And why did no one ever solve his murder? I teamed up with Peter to find out. With freedom-of-information requests, old newspaper archives, and interviews with retired detectives, he started to piece together the story of a person the system had forgotten.