In 2017, two things happened in advertising that I think about almost every week, though I only worked on one of them.
The one I didn't work on was the Pepsi commercial with Kendall Jenner. You remember it. A protest of indeterminate cause, beautiful young people of carefully calibrated demographics, a line of riot police, and Kendall, pulled from a photo shoot by some telepathy the ad never explained, handing a cop a can of Pepsi. The cop drinks it. The protesters cheer. Something has been resolved, though it's never clear what. The ad was pulled in forty-eight hours. Pepsi apologized. Nobody could be fired for it, because the thing that made it was the machine itself.
The other thing that happened in 2017 nobody remembers, because it was small and it worked. It was a short film called Inroads to the Community, and I wrote it. I was a junior creative at VIRTUE, VICE's in-house agency. My creative director, Tyler Pierce, let juniors run with ideas if they were any good, didn't steal credit, and didn't flinch when an idea meant telling a client something they didn't want to hear.
The client was Michelin. The brief came to us as a single declarative sentence I have never forgotten:
Show how tires = care.
I want to stay with that sentence for a second, because it's perfect. It's the entire condition of contemporary brand marketing compressed into four words and a math operator. A tire is a physical object that grips a road. Care is an orientation of the human heart toward someone else's wellbeing. The equals sign is doing more work than any equals sign in the history of semiotics. What Michelin was asking for was this: make people feel that our rubber donuts are actually a form of love, and do it in a way we can put on a reel.
Think about it for more than four seconds and it's an impossible brief. It's also the brief every brand is running right now. Show how oil = stewardship. Show how fast fashion = self-expression. Show how the algorithm = connection. The equals sign is always there, it's always lying, everyone in the room knows it's lying, and the job they actually hire you for is making the lie beautiful enough that nobody has to say so out loud.
I've taken to calling this condition the Pretend Era. It isn't a moral failure of individuals. It's a rational equilibrium. Everyone inside it knows the game is broken, and nobody can be the first to say so, because the first defector bears the whole cost of defecting. So the meetings continue, the briefs get written, the equals signs multiply, Pepsi happens, and then another Pepsi happens.
There's a way out, and it showed itself to me in that Michelin meeting. I've spent the nine years since trying to work out the rule behind it, so it could happen again on purpose. But first, the thing I knew walking into that meeting that Michelin didn't know about itself.
Most people under thirty-five have no idea that the Michelin Guide, the red book and the stars and the whole global apparatus of culinary legitimacy, is published by the same company that makes the tires on their Honda Civic. The Michelin brothers invented the guide in 1900 to give early French motorists a reason to drive farther, which would wear out their tires faster, which would sell more tires. It may be the most elegant piece of content marketing in history, and most of the people who care about it today don't know it has anything to do with rubber.
Michelin, I slowly realized, didn't know this either. They knew it institutionally, the way you know your own middle name, but they'd stopped feeling it. They'd been lost in brand-world sauce long enough that the guide and the tire company felt like two separate companies, and nobody in the room was twenty-four or lived in Brooklyn, so nobody had thought to ask.
I asked. I said, more or less, I don't think young people know the guide is you. The room went through the specific silence that happens when an emperor hears about his new clothes from a junior creative who probably shouldn't have been in the room in the first place. It was the silence of recognition: something everyone had half-seen for years, finally named, and impossible to unname.
That was the moment the brief changed, though nobody would have said so. On paper it was still show how tires = care. In the room, from then on, it was: how do we become legible again to the people who are supposed to care about us? No equals sign was going to answer that. The answer would come from showing up somewhere the brand had no obvious right to be, earning the right by doing something useful, and letting the usefulness be the message, because usefulness was the only message that couldn't be faked.
Which is where the chef comes in.
I interviewed a few chefs before I landed on Carlo Mirarchi. Carlo runs Roberta's in Bushwick and the tasting-menu restaurant Blanca, which holds two Michelin stars and serves about twelve people a night in a room behind the pizza ovens. I picked Carlo because he said yes before I'd finished explaining. Other chefs said yes the careful way chefs say yes to things that might eat a weekend. Carlo said yes like he'd been waiting for someone to call and ask, and I've learned to trust that kind of yes above almost anything in creative work.
The idea was to bring Michelin-starred chefs to the Grand Central Food Program, which is the largest mobile soup kitchen in the world and which I'm going to assume you've never heard of. Its vans run out of the Coalition for the Homeless and serve hot meals to unhoused and food-insecure New Yorkers along routes that have run nightly for decades. Hundreds of meals a night, every night.
I didn't want a gala or a fundraiser with a step-and-repeat. I wanted the chefs cooking the actual food, for the actual vans, on an ordinary night. And the film wouldn't be about Michelin. It would be about the vans, the chefs, the food and the people eating it. Michelin would appear the way a good host appears at a dinner party: visible enough that you know whose house you're in, invisible enough that you aren't thinking about them while you eat.
When I pitched it, the senior client from Michelin North America (I genuinely can't remember his name, and he deserves better than my faulty memory) said the thing that has stayed with me longer than anything else anyone said on the project:
“We don't feel we have the right to speak in this space.”
That sentence is the whole game. It's why Michelin came to VICE in the first place: they knew they needed to be culturally legible, and they knew the old Michelin Man campaigns weren't going to get them there. It's also why they'd been fighting the idea since I put it on the table: they were afraid of looking like they were renting moral weight they hadn't paid for. He was being careful rather than defensive, and I've learned to respect the difference.
What I said to him was something like: you earn the right to speak in this space by showing up and solving problems, by being useful rather than by having a position. And if you're worried the Michelin star is too luxurious to bring into a soup kitchen, consider that Massimo Bottura cooks for refugees and the unhoused through Food for Soul. José Andrés has fed more people in disaster zones than most governments. Nobody thinks their stars got smaller because they used them to feed hungry people. If anything the stars got bigger. The star isn't a fragile object.
He thought about it and said okay. It was a cautious, we'll see how this goes okay, the kind that means I'm giving you enough rope to hang yourself or to show me something I haven't seen, and I honestly don't know which. For my money, that's the most honest okay a senior client can give a junior creative.
The shoot happened on a night I remember as cold, though it probably wasn't that cold; memory exaggerates weather when it's trying to tell you something else. Carlo had come earlier in the day, and this is the part that took me a while to appreciate, not to perform chef-ness for the camera but to stand next to the GCFP's regular kitchen crew while they cooked. Most of that crew aren't formally trained cooks, and they make hot meals in batches that would break a restaurant. Carlo gave notes. He tasted things and suggested a little more salt here, a different finish there. The food that went out that night was theirs, a little better for having a two-star chef beside them for a few hours. I've come to believe that's the most accurate picture of what expertise is for that I've ever seen.
That night the vans ran their normal routes. The food went out the back doors, pre-plated, to people who had come to the stops the vans always go to, at the times the vans always come. Nothing was staged for us. We were joining the work.
The senior client showed up in person. He didn't have to. Senior clients at legacy industrial companies don't usually turn up in Harlem at nine at night for a shoot they've doubted for three months, but he came, in the kind of expensive Patagonia-type jacket corporate men wear to signal they're ready for weather, which is its own small form of care.
I watched him watch it. He stood off to the side while the crew handed plates out the back of a van to people who had come because they were hungry and the van was where the van was supposed to be. At some point he turned to me and said:
“I didn't understand why this might work until today. Now I get it.”
I've thought about that sentence more than anything anyone has ever said to me about my work. Not because it was flattering (it wasn't, particularly) but because it was honest, which is rarer and more useful, and because of the phrase until today. Three months of meetings and explaining and arguments about cultural capital hadn't produced that understanding. Standing on a cold sidewalk in Harlem watching food go out the back of a van did. It wasn't available in the conference room, only at the site of the thing itself.
It's the most important thing I learned in my twenties about how institutional change actually happens: you can't explain your way to it. You can only stand next to someone while it happens to them. The meeting is for permission. The site is for understanding. Confusing the two is the most common mistake made by people who want to change how large organizations behave.
After that night, I don't remember him fighting the project again.
The film aired. It was fine. It got a reasonable amount of attention for a reasonable amount of time and then receded into the archive of things brands have made about themselves. If the story ended there, you wouldn't be reading this. It ends, or rather begins, with the tires.
A few months after the film aired, the director of the GCFP emailed me. He asked if I'd go back to Michelin and ask for a van. An entire new van, so the GCFP could open a new route in the Bronx.
I wrote the ask. I argued that the project had shown what Michelin's capital could do when it stopped buying the appearance of care and started buying the infrastructure of care. I argued that a van in the Bronx was a better investment than any ad unit they could place with the same budget, because the van would still be feeding people in five years, in ten, in twenty, and no ad unit ever made has that kind of half-life. And I wrote in the part nobody had asked for, which is probably the thing I'm proudest of in a career that has since included a lot of things I'm proud of: Michelin would put its own tires on it. Tires are expensive, the GCFP runs on margins that would make an accountant weep, and Michelin makes tires.
Michelin said yes. They bought the van and donated the tires it rolls on. The GCFP opened the Bronx route, and it has been feeding hundreds of people a night, every night, ever since. I checked with the director this year: nine years on, the van is still on that route.
Notice what that does to the brief. The film showed chefs cooking food, which is a metaphor for care. But the tires, the actual rubber donuts, were now literally carrying the food to the people who needed it. In this one context, in the most literal way, tires equaled care. The metaphor collapsed into the thing itself, and the equals sign stopped lying.
That's the part that took me nine years to fully understand. The project didn't work because we made a good film; plenty of films are good, and most don't produce vans. It worked because somewhere in the process, the brand's product and the brand's stated value got physically connected through the body of someone who needed help. The tire was doing care, in the form of not going flat, which kept the van running, which kept the food arriving. For one brief moment the marketing and the material world were the same thing. A tire company signed a purchase order in 2017, and the vehicle did what vehicles do when they're looked after, which is keep going.
I didn't win a Cannes Lion for this project. I didn't win anything. If you search for Inroads to the Community Michelin 2017 you'll find roughly what you deserve to find, which isn't much. And yet there's a route in the Bronx that exists because of a meeting I sat in as a junior creative. Someone on it ate dinner tonight, someone will eat dinner on it tomorrow, and that's the only metric for creative work that has ever made sense to me.
It took me the rest of the decade to figure out why.
The thing I was slow to understand was myself.
For most of my twenties and the start of my thirties, people kept telling me, some kindly and some not, that I was too something for whatever room I was in. Too activist for journalism. Too journalist for brand work. Too media-theory for the people who had to ship by Friday, and too operational for the people who wrote about media theory in journals that came out twice a year. I'd gone to Duke for public policy, drifted into documentary work at VICE (an intern doing associate-producer work on the Emmy-winning HBO team, then the pre-launch VICE News crew), and then into advertising at VIRTUE, where I spent most of a decade as a story researcher, junior creative, strategist and insights publisher.
Every person who told me I was too something for the room wasn't wrong about me. They were wrong about the category. They assumed a world where journalism, brand work and media theory were separate disciplines with separate rooms, and in that world I really was a bad fit for all three. What I couldn't see, because I was busy trying to squeeze into one room and prove I belonged, was that the separation between the rooms was the broken part. The rooms were built for a world that was already gone, and nobody had noticed, because everyone was busy defending the walls of their own room instead of looking at the hallway.
The hallway, in 2017, was called VIRTUE. It was where the journalism instincts met the brand budgets met the media-theory brain, and, though nobody likes to admit it, it was the part of the building paying the rent while the rooms argued about who owned it. The Michelin project happened in the hallway. It couldn't have happened in any of the rooms. A journalism room would have rejected the sponsor, a brand room would have rejected the rigor, and a media-theory room would have rejected the deadline. The hallway had no walls, so it couldn't reject anything on principle. It could only ask whether the thing worked.
I spent the next few years figuring out what the hallway was. I had help: a pandemic.
In the spring of 2020, the world ran a collective stress test on its institutions, and they turned out to be unready in the ways that mattered most. Watching crisis-response agencies try to reach a population whose information environment had changed completely since 2005, I noticed that the most junior brand strategist at VIRTUE, earning maybe a fifth of what a senior FEMA communications officer earned, had a more accurate model of how information actually moves than the institutions whose job that was. The FEMA officer was almost certainly smarter and better credentialed. But the strategist had spent ten years inside the real information environment, watching which stories traveled and which messengers people trusted, while the officer had been trained to write the correct message and send it through official channels, a model that had been obsolete for about fifteen years. Nobody inside could say so, because saying so meant dismantling the premise that kept everyone employed. You can't see the water you're swimming in.
That was the moment I stopped trying to squeeze into rooms.
I built a company and called it The Auxiliary, because the word means the thing that helps the main engine run when the main engine can't run on its own. I'd become convinced the main engines of public-interest work weren't coming back the way they had been, and someone needed to build the thing that ran alongside them. The Auxiliary existed to do on purpose what the Michelin project did by accident: sit in the hallway between disciplines and produce outcomes none of the disciplines could produce alone. Aux Labs is what it grew into.
A room built in the hallway.
One more thing I want to name, and then I'll land this.
That condition, where everyone knows the pitch deck is nonsense and nobody can be the first to say so, is what I mean by the Pretend Era. I didn't invent the dynamic. Timur Kuran wrote the canonical book on it in 1995 and called it preference falsification. What I've noticed is that it has spread out of politics into nearly every institution I've worked inside, and seen through that lens, most of the last decade's public-interest failures look less like incompetence and more like an equilibrium nobody wanted and nobody could leave alone.
The way out of an equilibrium nobody wants isn't more exhortation; exhortation is what got us here. The way out is mechanism: a set of rules that changes the payoffs, so the thing everyone privately wants becomes the thing it's individually rational to do in public. That's what I've been building. The larger framework is the Trust-Repair Cascade, a ten-step sequence for how trust gets built and spent in public life. I won't walk through all ten steps here. I'll tell you the one that matters for the argument I started with: the step where brands, sponsors and holders of concentrated private capital come in.
The money to build what comes next sits in private hands that a quarter-century of CSR theater has trained to spend on performances of care instead of the material conditions of care. The oligarchs already own the brands, the media, the ad agencies and the foundations, and it's in their long-term interest that things keep working. Rory Sutherland has written well about the short-termism that stops them acting on that. I'm writing about the mechanism that would let them act without destroying the thing they're trying to help.
The answer CSR has given us is for the brand to commission the work, shape it and approve it, and to end up with a piece of content that rings hollow to every viewer, because the hollowness is built into the commissioning relationship. Pepsi in 2017 was the wrong answer. Gillette in 2019 was the wrong answer, preachy and well-meant and instantly rejected. Bud Light in 2023 was the wrong answer in both directions at once, which only a truly broken mechanism can manage: a brand punished by one audience for a gesture that did nothing meaningful for the audience it was aimed at, because the gesture was never connected to any infrastructure that could have helped anyone. The brand rented the appearance of allyship without building any of the pipes it would have needed to become real, and everyone could feel it.
Here's the rule I learned from the Michelin project and have spent nine years trying to formalize:
Brands can't be allowed to commission public-interest work. They can only amplify work that has already been independently ideated.
The research, the listening, the truth-finding and the artistic conception all have to happen in a space the sponsor can't touch, funded by institutions whose job is funding exactly that: foundations, universities, public-health bodies, philanthropies that still remember what they were for. The sponsor enters late, at the amplification stage, and what it buys isn't editorial control. It buys proximity to legitimacy that already exists, legitimacy it could never have manufactured on its own, because manufacturing it required the sponsor's absence.
That sounds like a restriction, but it works more like a gift, because it routes each kind of money to the stage where that money is productive. Foundation capital is good at funding diagnosis, ideation and rigor, because that's what it was invented to fund. Brand capital is good at funding scale, distribution and cultural reach, because that's what it was invented to fund. The firewall between the two stages isn't there to keep brands out of good work. It's there to make sure that when the brand finally arrives, the thing it's amplifying is still worth amplifying. Under CSR, brands pay retail for something that can only exist at wholesale, and they're baffled when it doesn't work. Under the firewall, the brand pays wholesale and the work works, because it was allowed to exist before the brand arrived.
The Michelin project was an accidental proof of concept. The GCFP had existed for decades; Michelin didn't invent it and couldn't have. What Michelin did, what a junior creative who didn't know he was building a mechanism got them to do, was show up late, after the work was already good, and help it travel further than it could on its own: a new van for a Bronx route on Michelin's own tires, and a film that pointed attention at the work without pretending the brand had built it. The brand didn't commission the care. It amplified care that was already happening, and earned the association it had been trying to get through the wrong door for years.
So here is the question I want to leave you with, if you run a brand, sit on a foundation board, or are the person in one of those rooms who has started to suspect the rooms themselves are the problem: what would it look like if the only thing your money was allowed to do was help work that was already good travel further than it could on its own, without commissioning it, shaping it or steering it? Would that feel like a loss of control, or like a relief?
There's a route in the Bronx that has been feeding people every night for nine years because somebody once said yes to the second version of that question without realizing they'd been asked it.
I'm still thinking about his face.