They drilled a hole in my skull on the 43rd floor of an empty skyscraper in Lower Manhattan. One of those towers where they told people to go and work from home and they never came back. Floor-to-ceiling windows, beige and white walls, spaces that felt impossibly big now that the cubicle dividers have vanished. One of those places where somebody pays to keep the lights on all night, every night, desperately trying to convince the world outside that this all still matters.
When I googled the place, I got the usual stories of illicit high-rise raves, the usual lifestyle content from the usual young influencers—sorry, creators: photos of them dancing, wide-eyed and ecstatic, rich off selling their own lives, silhouetted against the dawn sun creeping up over the Brooklyn skyline. But when I got inside, it was just startups squatting in random corners, and a bored security guard who scanned my face and the temperature of my skin before silently pointing me toward the elevator.
They drilled the hole in my skull in a medical tent squeezed into what used to be a server room. There were empty racks still bolted to the walls. They walked me in and laid me down, face first, on what I’m pretty sure was just a massage table, and fed me into a surgical robot that looked like a giant sewing machine. There wasn’t much else in there: a couple of laptops hooked up to a big touchscreen, an industrial air scrubber. Everything was plumbed into a mess of power and network cables that disappeared up into square black holes left by displaced polystyrene ceiling tiles. I only saw this briefly, but I remember it, because I remember being surprised there was no roof to the medical tent. It was literally a façade.
The last thing I saw was the scuffed vinyl floor, through that hole in the massage table, as they gave me the anesthetic and one of the technicians counted down from 10. I didn’t make it past 6. Then I was awake again, sitting up in a wheelchair looking out over that influencer view of Brooklyn, as they shined an LED light in my eyes and checked my reflexes. They gave me a Gatorade and a Kind bar and made me read some news stories out loud from an iPad while they squinted at my reactions on the laptop. A couple of hours later they made me get up and walk around, and as soon as they were happy I wasn’t going to have a seizure they gave me a blister pack of antibiotics and put me in an Uber home, which made me laugh. They paid for it, obviously. There’s no way they would have got the irony.
As if on cue, my phone decided I was hungry. It pushed a map of the surrounding neighborhood to the front of the screen. Places I didn’t recognize, or at least hadn’t eaten in for a couple of years, were highlighted with slowly pulsating pale blue dots. Two Chipotles, a gyro truck, some taco place I had a vague memory of, a soup and sandwich joint that actually looked new, and of course a Whole Foods Go. All of them were Amazon affiliates. I sat in the front of the truck and thought about how I was going to have to learn the city all over again, build myself up a new mental map of places I could easily go, lunches I could afford to eat. Or maybe I didn’t need to bother. Maybe I could just let my phone and the data brokers take care of it all.
A week earlier, my phone would have recommended the diner across the road, the one I was staring numbly at through the windshield. It would have known I was heading for it as soon as I turned onto this block, and then probably tried to distract me with coupon codes for some other, competing Uber-affiliated place. Now, it knew better. It knew going in there would be futile and painful and awkward. And, most important, it knew that going in there would be expensive. I didn’t belong in there anymore. My phone was coming to terms with my changing life a lot easier than I was.
It throbbed again, pushed something else to its screen. A text from Nakisha: where r u? everyone’s here! I started to compose a reply but stopped myself. It was pointless—I had nothing to say. Plus, if I did, it’d be weird to text it, when I could just cross the road and walk into the diner and say it to her face. She was in there right now, with a bunch of other Uber drivers. We used to try to come here at least once a month, just to catch up and hang, chat shit and bitch about how fucked everything was. If I was honest, it was always a little tough for me—being around people felt like a struggle at the best of times—but this was more about not being able to face any more goodbyes.
I stared at the diner some more, guessing who would be inside. I tried to hold their faces in my mind, but one by one they all slipped away from me, lost in the anonymous crowd. It was always there, a shuffling horde of familiar faces dissolved into vague, generic sketches. Friends, family, acquaintances, forever on the edge of my peripheral vision. Faces lost to me, swept away by the waves of sickness, death, change, and hard economics that had emptied out the city I once knew. Faces it was a whole lot easier to just let fade away than to try and find again, or to mourn.
Nakisha’s face didn’t fade, though, as hard as I tried to push it away, as much as I wanted it to be absorbed into the crowd. Let me make this clear now, though, to avoid any misunderstanding: there was never any hint of romance there. No unrequited this, no flirtatious that. There was no obsession, not from me. Just fear and awkwardness, and above all selfish guilt at not being able to accept genuine kindness and friendship, because I knew that one day they would also be ripped away. I hadn’t told anybody in there that I’d switched, not even Nakisha, and I was feeling pretty shitty about it. I’d met her when I’d started driving deliveries for Uber, just after Mayor Yang got elected. He’d swept in with a 1.7% lead over the other guy, which is what Americans call a landslide now, on a promise to solve New York City like a math problem. He’d promised everyone they’d get some money from the city every month to help us rebuild our lives and then, after the pandemic, from his Universal Basic Income experiment. But you can’t solve a math problem if you don’t know the numbers, and Yang didn’t get to see them until he was in City Hall, and then it was suddenly very clear that they didn’t quite add up. So he had to turn to the big tech companies—Facebook, Google, Uber, Amazon, the rest of them—to help him keep his promise.
So it was decided: you could take your measly token UBI payments from the city, or you could sign up with one of the tech giants and get a little more out of them. The big catch was that the companies didn’t even have to pay you in US dollars, so instead every month you got a little deposit of crypto into the wallet you had to install on your phone. Amazon Coin. ApplePay. FB Libra. Google Play Credits. It seemed complicated, but you only really needed to remember one thing: you’d better spend as much as your UBI with the company you signed up with—or its affiliates—because it went a lot further there. I mean, real numbers: we’re talking the kind of savings that not only meant you might make rent that month but made the poor old US dollar look worthless. Which was why I’d just switched—the owners of my apartment building had just flipped to being Amazon affiliates, which meant that if I continued paying with Uber Money I’d lose the discounts that meant I could afford to live there.
I stared at the diner, then back at the map dots on my phone, still pulsing blue as if they were trying to wake me up to the brutal truth. Forget this place. There’s nothing for you here. You don’t even have the right money. The people inside have no connection to you anymore. You’re not one of them. Time to move on.
I tapped its screen off, started the truck, and pulled away from the curb.
One week later and I was back on the abandoned 43rd floor, for a “calibration and orientation appointment.” Same server room, but the medical tent was gone, and instead I was sitting at a desk that looked like something hastily dragged in from some abandoned reception.
They’d given me the iPad again. It was playing a seemingly random montage of images and clips—news headlines, Beyoncé videos, Tom Brady winning yet another Super Bowl, influencers smiling in pristine kitchens, a cat knocking pens off a desk, a cheeseburger. America.
“Well, everything looks great,” the technician said, peering at me from over one of the laptops. “The implant seems to have taken root nicely. Getting a nice clear signal back—really clean, visible spikes.”
“Great,” I said, as though I understood her. “So what’s it look like in there?”
She almost laughed. “Well, I can’t tell you that exactly. But I can tell you’re probably pretty hungry right now, and you’re not much of a Patriots fan.”
She showed me a QR code, which I scanned with my phone, which made it install an app, which then showed me how to pair it with the implant. It was apparently no bigger than a grain of rice—she kept saying this, no bigger than a grain of rice, like it had been drummed into her by a marketing agency—and was wedged into the tiny hole the robot had drilled in my skull last week. Now it was sitting there, the skin grown back over the top of it, little hair-like sensors nudging the surface of my brain, waiting and watching for my dopamine levels to spike.
“There’s a lot of mythology around dopamine and what it does, and to be honest—like most things in the brain—we’re not completely sure how it works,” she told me, as I prodded around the app on my phone screen. “But put simply: by watching how and when it spikes, we can tell when you like something. Or at least we can tell when something makes you feel happy or fulfilled.”
“And that’s going to give me an edge with the data brokers?”
“That’s the plan, yeah. We record your dopamine levels. They get sent to Amazon alongside your usual everyday data footprint. Sync them up and we can see what’s really important to you, what makes you happy.” She sighed, sank back in her chair slightly. “These companies, Amazon and Facebook … they can take all the data they want now, but it doesn’t make it any easier to understand. They’re making assumptions about your behavior based on pattern matching and educated guesses. That’s really all machine learning is. But this—this is different. This is real. It’s actual correlation. It’s real insight. And that makes you—and your data—uniquely valuable. Hopefully we can get you a bit of extra coin.”
I stared past her and out through the windows. My head was full of thoughts: first the beautiful lifestyle influencers and high-rise raves, and then the possibility that I might make rent this month, and the app made my phone vibrate gently in my hand.
They told me to get out in the city, to enjoy myself. To see as much as possible. That’s why they’d been recruiting gig workers—drivers and deliverers: they needed a test population that was mobile, at a time when everyone else was still mainly working from home. They wanted to build a map of the new city, a guide to NYC’s greatest dopamine spikes.
The problem was I didn’t know which new city they meant. The Uber-affiliated city or the Amazon-affiliated one? The city of gig workers or the city of high-rise ravers? I guessed they meant the city that never sleeps, the city with a rough exterior but a heart of gold. But I was still stuck in the city where I’d heard nothing but ambulances and birdsong for four months, where the NYPD crushed spirits and skulls to remind us who was in charge, where we all sheltered in place, scared to leave our apartments. All I could see was the city where we let 50,000 people die and never paused to mourn them.
I didn’t know which city they meant, but I started to suspect I hated all of them.
For the first few days I was convinced the implant wasn’t working. Either that or it was my phone, an aging Samsung that had become even more sluggish after I’d switched. Now every third notification was an ad trying to get me to upgrade to the latest Kindle model, Amazon’s ecosystem reaching out to assimilate me even further. The app was meant to give me a notification every time I had a dopamine spike, but there was nothing, and I kept finding myself obsessively checking that the two were paired properly. Everything looked fine. It was supposedly calibrated so that it didn’t go off when routine activities gave me a minor hit—like taking a shit or dropping a package off on time. Maybe the calibration was off. Or maybe I just didn’t like anything anymore.
I put in a support ticket, but they got back to me saying the data feed looked fine, and I should just push myself a little harder—take some time to go find things I already knew I liked. Which I tried—even taking the kind of route deviations that pissed off Amazon’s driver management AIs just so I could check out some spot from the past. Some graffiti mural down in the Village, that one coffee place on Washington Square Park, the view back to Manhattan from DUMBO. I even tried wasting coins in non-affiliated food places I really couldn’t afford—cannoli and an espresso on Spring, falafel from Mamoun’s, a slice from Joe’s. Nothing. My phone sat there, unmoved, only buzzing to tell me I was running late for a drop-off, and that it was docking my payments.
And then it hit me, when I was sitting in the truck watching some kids shooting hoops and eating a too-expensive chopped cheese hero from a Facebook-affiliated bodega. I turned to the empty passenger seat to say something to somebody who wasn’t there.
I caught myself doing it over and over again. I’d reach out to touch their arm to bring their attention to something, or feel the muscle memory of taking their hand as we stared at the view. I’d eat a plate of food and load up a perfect bite on my fork to carefully feed to somebody, one hand gingerly hovering below to catch falling crumbs, so I could watch the smile spread over their face as they chewed it. I even found myself taking photos before realizing I had nobody to share them with beyond Amazon’s ad-tech algorithms or whoever might still be following my long-neglected Instagram feed.
The implant was fine; it was me that was unpaired. I’d fallen out of sync with the city, and I hated it for taking people away from me and leaving me on my own.
It’s safest to stay indoors: stay home. It’s the only way you can avoid the awkwardness, the disappointment, the fear. Delete Netflix and Uber Eats, install Prime Video and Amazon Restaurants. Stay at home and build your own city, make your own dopamine map. What’s the point of being lonely if you can’t do it by yourself?
At first, I tried to watch only movies set in New York, as though that had some significance. So the Avengers movies seemed a good place to start. I thought maybe watching the city being repeatedly reduced to rubble at a whim—endless computer-generated buildings demolished into nothing more than Technicolor pixel dust—might give me the hits I needed. But the app barely registered a spike for the first two hours and 22 minutes.
It wasn’t until I got to the post-credits scene—the one where the whole team is sitting around, silently eating, in some nameless, unaffiliated NYC shawarma joint—that my phone started to vibrate.
I’m not going to lie: for a fleeting moment I was ecstatic. I couldn’t tell you if it was just the dopamine spike or some joyful relief that the app had actually registered it. I rewound the scene and watched it again. Same spike, but with a slightly lower peak, according to the app. Third time was similar, but the results diminished again. Time to find more content.
At first I thought I’d have to watch whole movies for it to have the same impact—like I needed to build up some sense of connection with or investment in the characters before their friendships had any personal weight—and started earnestly slogging through the entirety of the Marvel Cinematic Universe. But a YouTube fluke showed me otherwise. Before I could stop it, autoplay served me a clip from Ant-Man and the Wasp where Ant-Man is playing with his daughter, and my phone vibrated in my lap. Repeatedly. It was a fucking revelation. I didn’t even have to sit through countless rubble-cities and the eternal melodrama and the endless wisecracking and the infinite polygons. Context was dead: all that mattered was fleeting, calculated emotional spikes.
It’s not hard to find the content once you know where to look. Listicles are your guide—the real maps to dopamine city are called things like “The 10 Most Heartwarming Moments in the MCU” or “The MCU’s 12 Best Friendships” or “Relive These Feel-Good Moments from the MCU.” Start by searching Tumblr and Screen Rant and you’ll find them all. It’s even better and more efficient if they give you the time stamps. Tony and James sniping at each other in Iron Man (00:10:42). Nick Fury buddy-copping with Carol in Captain Marvel (01:48:07). Peter Parker and Ned Leeds in whichever Spider-Man movie that was (00:23:38).
And then there’s the death scenes, which are perfect if you’ve also got some unresolved societal-level mourning to work through. When Killmonger dies in Black Panther. When Quicksilver dies in Age of Ultron. Spider-Man in Infinity War. Peggy in Winter Soldier. When Groot says “We are Groot” in Guardians of the Galaxy 2.
After a while, of course, you don’t need to search it out; it finds you. Before too long, every ad on every web page was screaming at me about young-adult-oriented TV shows I never knew existed and Star Wars spin-off cartoons. My YouTube recommendations filled up with nothing but fan-edited compilations of superheroes weeping, or supercuts of every time Frodo and Sam hugged.
There was this whole culture I’d avoided, that I thought I was somehow above, that wasn’t for me. An entire industry built to serve up comforting dopamine hits to a population wracked by technologically mediated loneliness, and exhausted by a society that felt like it was in constant, confusing collapse. People just like me, millions of us, PTSD victims refusing to mourn those dead or left behind, and resigned to being steamrolled by the decades of unease and disappointment still to come, as long as we could tap a screen to get a proxy emotion on demand, as if it were an Uber.
I’d been looping a fan edit of every furtive look between Anakin and Padme for four hours straight when the recall notification came through. I couldn’t ignore it—it paused everything else on my phone and splashed a message across the screen. Bright white text on blue, telling me how the implant was being deactivated immediately and explaining how I should contact the installers to have it removed. That was followed by a reminder that I was still under an NDA and that talking to the media about any of this could lead to legal action.
My first reaction was panic. Big “How the fuck can they take this away from me” energy. I’d just got this together. I finally had some structure, something that worked. My hand went to the implant scar on my skull—it was little more than a small bump now, but it felt hard at its center, and weirdly hot?
I slumped back on the couch, pulled up Gamora’s death from Infinity War on the TV, let it run, hoping it would calm me. But the anxiety didn’t fade away as usual. It seemed to dwell there, growing, behind my eyes and in the back of my jaw. I stared at the phone in my hand, waiting for the confirmation I needed, but it was lifeless, unmoving.
They took the implant out on the 43rd floor, in the same server room, in the same medical tent where they drilled a hole in my skull.
“Between you and me, it was always a fucking stupid idea,” the technician told me before she put me under. “You know how much work we had trying to set the filters so they didn’t register a spike whenever someone jerked off? We basically burned a whole round of funding on discovering that—guess what—people like porn.”
“And then there’s the people like you,” she said.
“Me people like mean what,” I said, fighting to keep my eyes open.
“People that get hooked into the feedback loop. Where they’re chasing spikes on the app, not what they actually find rewarding. I mean, it makes perfect sense when you think about it, bu—”
Then I was awake again, sitting in the wheelchair, staring out toward Brooklyn.
I sipped Gatorade and nibbled on a Kind bar while I watched them deposit enough Amazon Coin in my wallet to cover rent for at least four months. It was hush money, basically—payment for signing a waiver that said I would never sue them or talk about any of this to the media and rival affiliates.
They ordered me an Uber as soon as I felt I could walk, and on the way to the elevator I stopped and snapped a photo of the view on my phone. The city was laid out like a map in front of me, more inviting than I’d seen it in years. I was transfixed. Watching it felt as though a pressure had been lifted. There was so much still to do, but a start had been made.
Halfway across the Brooklyn Bridge, sunlight strobing through its towering structure, I reached for my phone and pulled my gaze away from the Uber’s windows. In silence I deleted the Prime Video and Disney apps, and wiped my YouTube history. I pulled up the photo from the 43rd floor and stared at it again. On some unexplained whim, I texted it to Nakisha.
Some minutes passed. My phone buzzed.
–whoooa, where the fuck are u?
–ha, long story. You wouldn’t believe me if I told you.
–try me. what about the diner, thursday?
My hand trembled. I forced myself to breathe, scared more of the screen than the city outside for the first time in as long as I could remember.
–sure, see you there
Geoffrey Hinton tells us why he’s now scared of the tech he helped build
“I have suddenly switched my views on whether these things are going to be more intelligent than us.”
Deep learning pioneer Geoffrey Hinton has quit Google
Hinton will be speaking at EmTech Digital on Wednesday.
Video: Geoffrey Hinton talks about the “existential threat” of AI
Watch Hinton speak with Will Douglas Heaven, MIT Technology Review’s senior editor for AI, at EmTech Digital.
Doctors have performed brain surgery on a fetus in one of the first operations of its kind
A baby girl who developed a life-threatening brain condition was successfully treated before she was born—and is now a healthy seven-week-old.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.