Stay informed
I Was Lonely and Let an App Pick My New Friends. Here’s How It Went
Last summer, I found myself on the market for new friends. I had just moved back home to Toronto after spending four years away for school. Many of the people I was closest to had scattered around the world for new jobs or degrees. It’s not in my nature to admit this sort of thing, but I was lonely.
Meta’s rapidly increasing mind-reading powers meant that, even before I really knew this about myself, Instagram had figured it out. I saw an ad for a platform that pitched itself as a way to make friends. I must have lingered on it for half a second too long. The ads multiplied. In them, attractive young women explained their reasons—unsentimental, unembarrassing reasons—for using such apps. Perhaps they’ve moved to a new city, or they love trying new things, or they live in San Francisco and only ever meet “tech bros.”
The ads were clear: these people aren’t desperate, and neither are you. They are young, fun, and hot. For a monthly fee—plus whatever you’d pay for dinner or drinks—you can off-load the labour of making lasting connections to an algorithm that will match you with a group you might click with. The model is catching on in big cities, like Los Angeles, New York, and more recently, Toronto.
I should be upfront: I was cynical about this experience from the start. I felt almost offended by the suggestion that friend-making could be outsourced to, or even facilitated by, an app. Dating apps exist, I reasoned with myself. But at least everyone knows they’re dehumanizing. These apps bill themselves as just the opposite: “We won’t tolerate an existence where we spend the majority of our time in a virtual world, robbed of our attention, meaning, & core life experiences,” one website declared. But what about the core experience of meeting people spontaneously, discovering the things that bind you, little by little?
These platforms are betting we no longer have the time, that we are too busy with our jobs, too addicted to our phones. Wouldn’t it just be so much easier to rely on them? I wanted to find out.
There is no shortage of apps promising friendship, each with its own theory of how connection should be engineered. Some sell exclusivity; others offer efficiency and scale. Some are quite general, while others, like Outclose and The Breakfast, offer paid friend-making experiences with individual twists (The Breakfast, for example, curates breakfasts).
I decided to sign up for 222, which began around early 2021 as a university-funded research project with the goal of predicting whether people would want to spend time together based on data about their identities, values, and beliefs. Eventually, the project transitioned into a real-world experiment in which co-founder Keyan Kazemian hosted backyard dinners at his house in Orange, California, with groups of algorithmically paired strangers. As the experiment grew, the founders secured funding from Y Combinator and made the leap into a fully fledged company in 2023.
When an invitation for dinner followed by drinks at an “intimate speakeasy” popped up on my screen, I felt a thrill. Chic web design, fancy-sounding evening out—it all felt so premium. And the curation was serious business. I had spent nearly fifteen minutes filling out an extensive questionnaire (other 222 users reported spending up to an hour), with questions ranging from “how extroverted are you?” (somewhat) to “rate your physical attractiveness from one to ten” (um, seven?) to “do you worry that other people don’t really love you?” (not really) to “do you believe that early term abortion is always morally permissible?” (yes).
The restaurant, a Portuguese place, was quiet. The six of us were given a name for the reservation and shown to a table, rather than having to awkwardly identify ourselves as users of a friend-making service. The bar that we relocated to after dinner (Mahjong Bar; extremely trendy) was only an eight-minute walk away, and an entire table was reserved to store our coats. It was clear someone had thought of everything in order to make the process as smooth as possible.
The algorithm announced itself from the moment all of us ordered the same drink. Kazemian told me this is what sets 222 apart from competitors: the curated matching. When the waiter brought out six identical pink cocktails, I believed him. The people at this table had been selected for me, and I for them. Just like the ads promised, they really were young, fun, and hot. As the night went on, the overlap deepened: reading, writing, movies. We were legible to each other. We knew what sorts of questions to ask. It was, overall, easy.
Are these things sufficient foundation for a friendship? Two people might check a box that they enjoy reading, but do they enjoy the same books? Do they enjoy reading for the same reasons? These strike me as more difficult things to quantify, but I might not be giving 222 enough credit. The algorithm’s most powerful feature, says Alexander Saiko, 222’s head of operations for Toronto, is its ability to learn the preferences of its users, matching them in ever more sophisticated ways. According to Saiko, it takes about five experiences for the system to adapt. Admittedly, I only attended one.
But 222’s user base might be an inherent limitation—at least for now. The platform selects for high earners because the app is expensive. The roughly $30-per-month subscription fee might not sound terrible (although, it’s more than I usually spend on making and keeping friends), but you also have to cover the cost of each experience. Naturally, the platform is going to select for people willing to splurge on a semi-regular basis. It also likely favours people whose long work hours crowd out hobbies. Our era of hybrid and remote work has created a cohort of young professionals with money but a thinning social circle. This seems like a clever demographic to target, but it’s also a narrow one. If you’re trying to find everyone their perfect friend, a diverse pool seems key.
Saiko didn’t seem concerned. “I don’t think there’s any average user,” he told me. “The bigger 222 gets, the more it becomes reflective of the population.” And at least in some cases, 222 works. I interviewed a group who met through the app; most of them were piled into a car on the way to a climbing gym when they joined my Google Meets call. They passed the phone around so that different people could answer my questions, laughing the whole time, sharing brief moments of eye contact that made their closeness unmistakable. Their feedback about the platform was glowing.
Don Xu told me he has attended upwards of thirty 222 experiences, and he praised the platform’s ability to offer unique activities, like boxing, crafting, or visiting a museum. Frank Nguyen stressed that the platform takes user feedback very seriously, seeming to add or remove future events based on shifting preferences. Alina Liu mentioned that the strict cancellation fees ensure people actually show up to the events. Everyone praised the algorithm. “It grouped me with very compatible people,” Xu told me.
If 222 is the designer brand of friendship apps, Timeleft is Walmart—a cheaper way to get what you’re after. And you get what you pay for. The personality quiz took all of two minutes, and when I arrived at The Berczy Tavern for dinner, some unspecified miscommunication between Timeleft and the restaurant meant we weren’t seated until forty-five minutes after our reservation. Confused staff first led me to the wrong table. Perhaps that explains why, when we finally did sit down, there were only three of us instead of the promised six.
My two companions had both used Timeleft before. One of them had just moved to Toronto from Vancouver and was using it to make new connections in the city. The other was something of a Timeleft regular; he invited me to join a WhatsApp group chat with over 400 members. Both companions were locked into multi-month subscription plans—one for three months, the other for six. The platform, they told me, is imperfect. One of them recounted being sent to a high-end sushi restaurant, despite selecting a preference for budget-friendly dining. His group left and went for tacos instead. Indeed, The Berczy Tavern was also a much nicer restaurant than I had anticipated. I ordered one of the cheaper entrees: a $36 truffle risotto.
Timeleft’s appeal did not seem to lie with the expert curation. Toward the end of the dinner, I got an email from the company inviting me to an “exclusive After Party” with all the groups in the city from that evening. The email said it was at Greta, an arcade bar on King West. My companion groaned. It’s always either Greta or Belfast, he told me, and Greta is the worst. Sometimes, he said, when Timeleft sends everyone to Greta, the WhatsApp group chat revolts and chooses another bar.
I came into the experience determined to do Timeleft right, to faithfully engage in every step. But it became clear to me that doing Timeleft right meant using the platform as a springboard rather than a place to land. The app offers a virtual deck of cards with questions for getting to know each other, but neither of my companions indicated any interest in looking at them. I followed their lead, and we talked about work, sports, and movies instead.
What Timeleft offers, it seems, is shockingly simple: dinner with strangers. No more, no less. If you’re new to a city, if you’re sick of only talking to people who work in your field, if you want to broaden your dating pool, it sits you down next to people you never would have run into otherwise and leaves the rest to fate. Many people I spoke to also pointed out how great these kinds of platforms can be for solo travellers who want to have a fun night out in a new city.
Reflecting on the city-wide group chat, the solo travellers, and groups of strangers fleeing fancy sushi restaurants for tacos, I thought about Donna Haraway, a theorist from the 1980s. In “A Cyborg Manifesto,” her most well-known work, she describes humans in a digitally mediated world as cyborgs, or “illegitimate offspring” of the systems that govern our lives. In essence, she argues that a technology’s radical potential does not lie in the execution of its creator’s vision—almost always a money-making one—but rather in how everyday people reimagine what (and who) the technology might be for.
Are Toronto’s Timeleft users cyborgs, engaging with the app in unexpected ways and creating new avenues for community building? I got added to the Toronto group chat to see what it was all about, and I’ve had messages flooding my phone about comedy shows, vintage markets, brunch spots, and pickleball ever since.
After dinner, I walked to Greta. Neither of my companions came to the bar, but one of them—the Vancouverite—told me he was headed that way anyway, and so we walked together. At dinner, I had found him polite and friendly, but I felt early on that we were unlikely to be friends. He had spoken excitedly about watching sports, and I had spoken excitedly about reading novels. For an algorithmic match, we had little in common.
But as we walked, our interactions started to loosen, feel less contrived. Instead of asking “what’s your favourite” questions, we started telling each other stories. We started laughing. We started talking about our lives, our identities. I remember startling myself with my own honesty.
He and I haven’t really been back in touch since then, and I’m not sure whether we ever will be. But I’ve started to wonder if the promise of the algorithm isn’t just false but counterproductive; if we think we’re going to dinner with people who have been data profiled for us, we’ll rule them out quickly when they don’t suit our expectations. We’ll chew awkwardly on our truffle risotto, grasp at more ways to talk about the weather, and decline the dessert menu so that we can get out of there faster.
But when we allow ourselves to fall into a moment—even knowing the fit is imperfect, even if the moment is just a twenty-minute walk down King Street—we might be more willing to give one another a chance.
Timeleft and 222 are the most dominant platforms in the friendship-curating space; 222 has over 150,000 Instagram followers, which is impressive except when compared to Timeleft’s 1.2 million. Both platforms insist that they got to the concept first, and when I spoke to company representatives, both suggested that other platforms were lower-quality rip-offs.
“We had a lot of copycats after we launched Timeleft,” said communications and partnerships manager Aya Tohme. But the other companies, she said, don’t have “a deep mission like us.” 222’s Kazemian offered a similar sentiment, saying that “this space has become filled with copycat clones” and that other platforms are “cheaper, faster products.”
The choices of words from Tohme and Kazemian are strikingly similar—as are, of course, the basic premises behind their companies. But there are other similarities, harder to explain away. On April 24, 2023, Timeleft co-founder Maxime Barbier posted on his Instagram account, announcing the launch of artificial-intelligence-powered social matching. “Not a dating app. Not another distraction. No Mindless scrolling. No profiles, no DMs. No scrolling or swiping,” the Instagram post reads.
It reads a lot like the description on 222’s webpage: “This is not mindless scrolling . . . This is not a distraction.” And, toward the bottom of the page, “no profiles, no DMs, no scrolling, no swiping.” According to a digital archive of the site, a version of this copy (which also included the phrase “This is not a dating app”) has existed on the 222 website since December 2022, months before Barbier announced his version. (Timeleft did not respond to The Walrus’s request for comment regarding these parallels.)
I asked Tohme outright if Timeleft had stolen 222’s idea. She denied this, and she said that Timeleft has been focused on facilitating human connection since its inception. Originally, Tohme told me, the platform was structured around people sharing their hopes and dreams. A digital archive of the company’s website from spring 2021 reads: “Dream big, start now. Create your bucket list and share it with thousands of people who can help you achieve it.”
By summer 2022, Timeleft had pivoted to offering “cool experiences with cool people” instead, with a dating-app-style interface for swiping on activities and other users. It transformed again around March of 2023, leaning more into the “cool people” side of its branding. And then, a month later, well after the launch of 222, it introduced a new feature, Timeleft IRL: dinner with strangers—no profiles, no DMs, no scrolling, no swiping. “Timeleft was always there, since five years ago,” Tohme told me. “So, I don’t think that we stole 222’s idea. Timeleft’s founders had the idea but just pivoted more and more and refined the concept.”
Timeleft’s Barbier has had his fair share of controversy. He faced multiple accusations of sexual harassment back in 2017, eventually apologizing for “heavy and persistent seduction.” He’s also been accused of plagiarism. Back in 2014, YouTuber Casey Neistat called him out for allegedly stealing a short film he had made two years prior. “I watch and draw inspiration from my favorite filmmakers,” Neistat wrote on Tumblr. “What this Maxime character did was different.” (TimeLeft also did not respond to The Walrus’s request for comment.)
Still, Barbier’s scandals don’t seem to have hurt the company; Timeleft currently operates in more than 200 cities around the world, and its website boasts over 3 million members.
The “offline wave,” as Groupvibe founder Khalid El Haji called it, has many hopeful startups trying to ride it. But even when relationships aren’t outright hostile, companies seem hesitant to acknowledge each other. “I have to be really careful here,” El Haji told me when I asked him about Groupvibe’s relationship to similar platforms. Collaborations between companies aren’t unheard of (Groupvibe is currently exploring a partnership with a similar platform), but the overall atmosphere feels strikingly tense. The stakes are high, after all: they’re fighting for a monopoly on community. Saiko told me that 222 aims to be the “default social facilitator” for its users. It doesn’t just envision itself as a friend-making service but as a friend-keeping tool and a romantic matchmaker. These apps want us to trust them with every facet of our social relationships—more than we trust ourselves.
And maybe that would be easier. Whereas generations past might have relied more on family and geographic community ties, writes Jennifer Senior for The Atlantic, “it’s precisely because of the atomized, customized nature of our lives that we rely on our friends so very much.” More than ever, friends are important to have and difficult to make; these apps can help with that.
But the fact that we get to choose our friends means we get to choose something about ourselves, a unique privilege in our cultural moment. Algorithms are getting harder to escape—Spotify shapes our music tastes, meanwhile Substack and Instagram shape our opinions on everything from fashion to politics. Our senses of self are being constructed, increasingly, by ads, clickbait, and AI-generated slop.
Friends influence our beliefs and preferences from entirely different angles, yanking us out of our digital echo chambers and tethering us to the real world. What happens when they become algorithmically curated too?
“Market forces are pushing us in a way that’s trying to strip us of our agency,” Saiko noted, echoing many of my own concerns. But he positioned 222 and its algorithm as firmly on the side of the people. “We think the technologies that exist can be pointed toward humanity,” he told me, “instead of pointing toward anti-human purposes.” I thought of Haraway again. “The cyborg would not recognize the Garden of Eden; it is not made of mud and cannot dream of returning to dust,” she writes. In other words: well, we’re here now, probably already more machine than human. Can these platforms help us make the best of that?
I don’t know. Amidst eight-hour daily screen times and an uptick in people hitting it off with chatbots, these platforms have led to lasting human friendships; I think that’s a good thing. At the same time, as these platforms try to remove every possible challenge associated with human connection, I can’t help but wonder if they’re removing the very things that give our relationships meaning: the risk and reward of “putting ourselves out there,” of meeting each other serendipitously and choosing each other on purpose.
I think most people have wished, at one time or another, for relationships to be easier. Making friends can be frustrating; you have to make time for a new person in your busy life, put energy into planning dinner dates when you might rather watch TV at home, and get comfortable being at least a little vulnerable with someone who may or may not reciprocate. Keeping friends can be even harder. Sometimes you have to pick up the phone when you’re in the middle of something, offer advice that should be painfully obvious, go see their new place even though they don’t live anywhere near the subway line anymore.
Timeleft and 222 are good at what apps are good at: reducing a human problem to an interface and price point. But, for me, the best kind of user experience—maybe the only kind I’m interested in—is the kind that feels worth some effort. Caring is not a UX issue. It is still, stubbornly, an art of wild inconvenience.
The post I Was Lonely and Let an App Pick My New Friends. Here’s How It Went first appeared on The Walrus.
Comments
Be the first to comment