Yes Theory: Running Out of Fuel Above the Mariana Trench Is What a Real Boundary Looks Like

Source: Colin and Samir | Published: 2026-09-23T13:13:32Z

Yes Theory spent $10,000 on a taxi joyride, but where they went didn't matter. Driver Reg turned out to be the real star of the episode.

He's the co-founder of Yes Theory. The channel has been running for 11 years. At 10,000 subscribers, he felt the same way he does today at 10 million: "I'm still at the beginning." Only 2,600 channels on YouTube have 10 million subscribers. Yes Theory is one of them — and one of the few still going after more than a decade.

What It Feels Like to Make Positive Content Inside a Negative Feed

Thomas has a phrase for what they've always been doing: "hiding the vegetables."

The story itself has to be human, has to carry real value — but the packaging needs tension. Titles and thumbnails need a conflict, a mystery, a question you can't help but click on. "50 Hours on Europe's Most Dangerous Island" — he admits that's a stretch, but the island did have genuine safety concerns, and when they got there, they built real connections with the local community.

The vegetables are real. The packaging exists to make you want to eat them.

He says he always assumes nobody cares about any given video, and starts from there: How do I make someone care about this? Skydiving videos went nowhere for a long time. Then they reframed it as "pulling random strangers off the street and jumping right now" — suddenly it was a concept. Same content, different frame, completely different result.

A Dream That Changed the Channel

In 2020, the pandemic killed their entire production model. The core format — dragging strangers into adventures — became impossible overnight. Sitting at home, Thomas started questioning the format itself: was it starting to feel mechanical? They knew exactly how every segment would go. That, he realized, was the problem.

During that period, he kept having the same dream — he was sitting in Afghanistan, drinking tea with a man he didn't know. He mentioned it to Drew Binsky, a fellow YouTuber who happened to have a trip to Afghanistan planned a month out. Thomas went with him.

He describes the trip as getting his mind "cracked open." From then on, Yes Theory's approach shifted. No more preset payoff moments, no tidy narrative arcs — just showing up somewhere with open curiosity, following the energy, seeing who you meet. They now deliberately under-plan before departing. That under-planning forces everyone to stay fully present. When they encounter someone genuinely worth following, the whole team locks in. Sometimes a three-to-four-hour chance meeting becomes the entire episode.

Three Ways Creators Think Differently from Filmmakers

Thomas and his host discussed a question both had been sitting with: as YouTube and Hollywood converge, what exactly is "creator thinking" — and how is it different from traditional film?

The host distilled it to three things.

First, platform context awareness. Traditional filmmaking is deliberately medium-agnostic — a film should work in theaters, on TV, across screens. YouTube creators are building for a specific environment. They know where the audience is sitting, when they're watching, and what happens the moment the video ends.

Second, empathy for attention. On YouTube, the next video is always one click away. You have no claim on anyone's time — you have to earn every second they stay. That awareness bleeds into every filming and editing decision.

Third, collaborative intimacy with the audience. YouTube viewers don't want to be passive recipients. They want to feel like they're part of something. When a creator says "last time you mentioned..." or "we did what you suggested" — that creates a bond no other platform can replicate.

Thomas added: they never pretend the camera isn't there. The camera operator appears in frame, speaks, gets introduced to the people being filmed. That's not a flaw — it's honesty. "You can't fool the audience. Especially an audience that grew up on YouTube."

Finding Stories Nobody Has Told Yet

Over the past year, Yes Theory published 25 videos and pulled 75 million views.

Thomas thinks about new ideas every day. He's heard so many pitches that his bar has gotten high — it has to be something he's genuinely never seen, or an angle nobody has pulled off before.

He spends real time randomly wandering Google Maps: zoom in, keep zooming. What is this island? Click, search, discover it's a remote research station, start trying to make contact. Tokyo's underground drift scene was supposed to be inaccessible — then a French YouTuber friend, Simon Puechuech, found a way in, and it became an episode. Switzerland's 374,142 bunkers: he was struck by the sheer absurdity of the number. The world's most famously neutral country has assigned every single citizen a bunker number. If things go wrong, you call in and find out which mountain to run to. They went. They found that some WWII-era military bunkers had become too expensive to maintain and were sold off — now wealthy people rent out rooms in them as emergency refuges.

He mentioned one project he can't talk about yet — a place where even getting there is unsolved, potentially requiring a six-figure logistics budget. "If we can figure out how to get there, we can tell a story nobody has ever told."

The Man Living Alone on That Island

The Northern Mariana Islands are U.S. territory. Almost no Americans know they exist. While traveling there, they met a subscriber who told them: you have to go to Pagan Island — there's supposedly one person living there alone.

Pagan has an active volcano and is nearly impossible to reach. They chartered a fishing boat. Midway there, they ran out of fuel and drifted — directly above the Mariana Trench, the deepest point in any ocean on Earth. "In that moment we had genuinely reached the edge. One more step and it would've been the unknown."

They made it. They found the man. His name was Jordan. The first night, they put the cameras down and just talked — hours of it. Jordan was guarded at first, hard to read. But by the end of the night, both sides had relaxed. The next morning he woke up visibly different. Like a different person.

Thomas says they kept exchanging glances: "The things he's saying right now are incredible. We should be recording." But that night wasn't for content. It was for actually becoming friends. "If we had been recording, we wouldn't have had the next day."

Spending $10,000 on a Taxi Ride

Even they thought this one sounded a little ridiculous.

The idea: hail one of London's iconic black cabs, let the meter run to ten thousand dollars, see where you end up. It's YouTube-native, visually distinctive, and it conveniently called for a native English-speaking driver.

Thomas stepped out of a train station before filming had even started and flagged a random cab. The driver was Reg — a British-Indian Sikh taxi driver in a turban, speaking in a thick East London accent. "I had never seen that combination in my life." Reg had made an illegal U-turn on a one-way street just to pick him up.

They spent far more time together than any single video could hold, and it eventually became a series. Thomas and Reg have now known each other for a long time. When they met, Reg's life wasn't easy — he was carrying debt — but he showed up every day, and treated everyone who climbed into his cab with the same warmth. Thomas said: "The ten-thousand-dollar ride was just an excuse to sit next to him. Where we ended up didn't matter."

"The concept is a tool for meeting people. It was never the point itself."

In the end, he says, what you remember isn't subscriber counts or analytics. It's Reg. It's Jordan. It's the real moments, like training with Stefan in Finland at negative 35 degrees Celsius. Stefan got frostbite on several toes on that trip — they went numb for months. They could have done part of the experience and retreated somewhere warm to sleep. The Finnish military explicitly offered that option. They said no.

When AI Starts Mass-Producing "Formatted" Content

Thomas says the YouTube channels he loves most right now are the ones where you can feel it — this person is doing what they love. Magnus the Norwegian climber. Mike Corey. No big team, no heavy optimization. But the stories are real. The people are real.

His read: AI will get increasingly good at producing highly structured, highly polished content — and that kind of content will keep getting cheaper. At the same time, people will grow increasingly hungry to feel a real human on the other side of the screen.

He's not saying optimization is useless. He's talking about priority. If optimization becomes the first thing, you lose the reason you started. He treats optimization as a toolbox — but the starting point is always: Am I actually curious about this?

Josh Richards, one of TikTok's early breakout stars, was asked at an event what advice he'd give creators. He said something Thomas found unexpectedly profound: "If you're not having fun making it, how do you expect the audience to have fun watching it?"

Thomas's honest reaction in the moment was mild condescension: What does Josh Richards have to say? But the line stuck.

The Internet Numbs You — Which Makes Realness More Valuable

He described a moment he witnessed firsthand: in a Venice Beach studio, news broke on Instagram and TikTok that someone had been stabbed nearby. Everyone pulled out their phones. Five minutes later, everyone went back to work.

"We were quietly experiencing something very strange together — and we weren't talking about it."

That feeling, he says, is part of what drives him to create: the desire to preserve something real inside all that numbness. When filming, his goal is to make the people in front of the camera feel safe enough to be vulnerable. That usually means building a relationship first — spending hours with someone before the camera ever turns on. They'll often wait until the very end of a shoot to mention this is a channel with 10 million subscribers — not to deceive anyone, but to keep the number from changing how they behave.

He has one test: after a video goes out, if people from that community show up in the comments saying "we were seen" — it worked. If anyone feels consumed or misrepresented, it doesn't matter what the data shows. It failed.

"I'm a bridge. The story doesn't belong to me — it belongs to that person. My job is to carry them, as faithfully as I can, in front of an audience."