IN ABOUT 40 minutes, Cindy Manit will let a complete stranger into her car. An app on her windshield-mounted iPhone will summon her to a corner in San Francisco鈥檚 South of Market neighborhood, where a russet-haired woman in an orange raincoat and coffee-colored boots will slip into the front seat of her immaculate 2006 Mazda3 hatchback and ask for a ride to the airport. Manit has picked up hundreds of random people like this. Once she took a fare all the way across the Golden Gate Bridge to Sausalito. Another time she drove a clown to a Cirque du Soleil after-party.

鈥淧eople might think I鈥檓 a little too trusting,鈥 Manit says as she drives toward Potrero Hill, 鈥渂ut I don鈥檛 think so.鈥

Manit, a freelance yoga instructor and personal trainer, signed up in August 2012 as a driver for Lyft, the then-nascent ride-sharing company that lets anyone turn their car into an ad hoc taxi. Today the company has thousands of drivers, has raised $333 million in venture funding, and is considered one of the leading participants in the so-called sharing economy, in which businesses provide marketplaces for individuals to rent out their stuff or labor. Over the past few years, the sharing economy has matured from a fringe movement into a legitimate economic force, with companies like Airbnb and Uber the constant subject of IPO rumors. (One of these startups may well have filed an S-1 by the time you read this.) No less an authority than New York Times columnist Thomas Friedman has declared this the age of the sharing economy, which is 鈥減roducing both new entrepreneurs and a new concept of ownership

The sharing economy has come on so quickly and powerfully that regulators and economists are still grappling to understand its impact. But one consequence is already clear: Many of these companies have us engaging in behaviors that would have seemed unthinkably foolhardy as recently as five years ago. We are hopping into strangers鈥 cars (Lyft, Sidecar, Uber), welcoming them into our spare rooms (Airbnb), dropping our dogs off at their houses (DogVacay, Rover), and eating food in their dining rooms (Feastly). We are letting them rent our cars (RelayRides, Getaround), our boats (Boatbound), our houses (HomeAway), and our power tools (Zilok). We are entrusting complete strangers with our most valuable possessions, our personal experiences鈥攁nd our very lives. In the process, we are entering a new era of Internet-enabled intimacy.

This is not just an economic breakthrough. It is a cultural one, enabled by a sophisticated series of mechanisms, algorithms, and finely calibrated systems of rewards and punishments. It鈥檚 a radical next step for the 颅person-to-person marketplace pioneered by eBay: a set of digi颅tal tools that enable and encourage us to trust our fellow human beings.

Manit is 30 years old but has the delicate frame of an adolescent. She wears a thin kelly-green hoodie and distressed blue jeans, and her cropped dark hair pokes out from under her purple stocking cap. Yet despite her seemingly vulnerable appearance, she says she has never felt threatened or uneasy while driving for Lyft. 鈥淚t鈥檚 not just some person off the street,鈥 she says, tooling under the 101 off-ramp and ticking off the ways in which driving for Lyft is different from picking up a random hitchhiker. Lyft riders must link their account to their Facebook profile; their photo pops up on Manit鈥檚 iPhone when they request a ride. Every rider has been rated by their previous Lyft drivers, so Manit can spot bad apples and avoid them. And they have to register with a credit card, so the ride is guaranteed to be paid for before they even get into her car. 鈥淚鈥檝e never done anything like this, where I pick up random people,鈥 Manit says, 鈥渂ut I鈥檝e gotten used to it.鈥

Then again, Manit has what academics call a low trust threshold. That is, she is predisposed to engage in behavior that other people might consider risky. 鈥淚 don鈥檛 want to live my life always guarding myself. I put it out there,鈥 she says. 鈥淏ut when I told my friends and family about it鈥攅ven my partner at the time鈥攖hey were like, uh, are you sure? This seems kind of creepy.鈥

That skepticism reflects a widely held, deeply ingrained attitude 颅reinforced by decades of warnings about poisoned Halloween candy and drink-颅spiking pickup artists. No wonder some of the loftier 颅sharing-颅economy executives see their mission as not just building a business but fundamentally rewiring our relationships with one another. Much as the traditional Internet helped strangers meet and communicate online, they say, the modern Internet can link individuals and communities in the physical world. 鈥淭he extent to which 颅people are connected to each other is lower than what humans need,鈥 NYU professor Arun Sundararajan says. 鈥淧art of the appeal of the sharing economy is helping to bridge that gap.鈥 Lyft cofounder John Zimmer goes so far as to liken it to time he spent on the Oglala Sioux reservation in Pine Ridge, South Dakota. 鈥淭heir sense of community, of connection to each other and to their land, made me feel more happy and alive than I鈥檝e ever felt before,鈥 he says. 鈥淚 think people are craving real human interaction鈥攊t鈥檚 like an instinct. We now have the opportunity to use technology to help us get there.鈥

But we鈥檙e not there quite yet. Data from the 2012 General Social Survey, the National Opinion Research Center鈥檚 poll of American attitudes, found that only 32聽percent of respondents agreed that people could generally be trusted, down from 46聽percent in 1972. More recently, an October 2013 AP-GfK poll of more than 1,200 Americans found that just 41聽percent of respondents express 鈥渁 great deal鈥 or 鈥渜uite a bit鈥 of trust in the people they hire to work in their home, only 30聽percent trust the cashiers who swipe their credit or debit card, and a mere 19聽percent trust 鈥減eople you meet when you are traveling away from home.鈥
She pauses, mulling it over for a few more seconds.Even Manit isn鈥檛 willing to fling open her doors to every sharing service that comes along. For instance, she isn鈥檛 all that comfortable with the idea of letting strangers rent her car, as she could through companies like RelayRides or Get颅around. 鈥淪omeone I don鈥檛 know taking my car鈥攖hat鈥檚 different,鈥 she says. 鈥淚 have to be there with them.鈥

鈥淲hat I鈥檇 wonder is, what are they doing with my car?鈥 She lets out a little laugh. 鈥淟ike, what are they doing with my car?鈥

On an unseasonably balmy mid-drought morning in January, I walk about 20 blocks from my home on the south side of San Francisco and knock on the door of a guy named Paolo, who promptly hands over the keys to his 2013 Subaru Impreza. Paolo drives his car only on weekends, so he鈥檚 free to rent it out through RelayRides the rest of the week. Paolo says he鈥檚 never had a second thought about letting a stranger drive off with his vehicle, perhaps because he is 鈥渦nreasonably trusting,鈥 as he describes himself. (Though not, apparently, trusting enough to let me publish his real name, which is not Paolo.)

I mostly avoid texting while driving Paolo鈥檚 car to Santa Clara, where I meet with Rob Chesnut. A former federal pros颅ecutor, Chesnut created the trust and safety department at eBay. Just as PayPal incubated a mafia of ambitious technologists and business leaders, a related eBay mafia has spread its tendrils throughout the sharing economy, with members at Airbnb, RelayRides, Task颅Rabbit, and oDesk. Chesnut himself is a senior vice president at Chegg, an online platform for college students, but he serves as an adviser to sharing companies like Elance and Poshmark.

Chesnut first came to eBay as a customer in 1997, in search of a Polaroid SX-70. This was early in eBay鈥檚 development, when its trust and safety policies could be summed up by founder Pierre Omidyar鈥檚 animating premise: 鈥淧eople are basically good.鈥 Chesnut did not find this particularly reassuring. 鈥淎s a federal prosecutor, I鈥檓 not an altogether trusting type,鈥 he says. 鈥淚鈥檓 used to dealing with the worst of society all the time. Now I鈥檓 going to send a cashier鈥檚 check to a total stranger?鈥

Of course, we engage in commerce with total strangers every day. We hand our credit cards to shop clerks, get into the backseat of taxis driven by cabbies we鈥檝e never met, ingest food prepared in closed kitchens, and ignore the fact that hotel workers with master keys could sneak into our rooms while we sleep. But each of those transactions is undergirded and supported by a complicated series of regulations, backstops, and assurances that go back to the Industrial Revolution.

Before that time, Americans tended to cluster in small towns and farming communities, where citizens built tight-knit relationships over the course of many years. In an economic system like that, where everybody knows everybody else, there鈥檚 a natural incentive to treat people well: Get a bad reputation and the whole town will know about it. On a broader level, the members of these small, homogeneous communities knew that their neighbors probably saw the world in the same way they did, holding the same morals and belief systems, which made it easier to conduct business with them.

That all started to change around the mid鈥19th century. As Ameri颅cans moved from small towns to big cities, small merchants were replaced by large corporations, and local markets gave way to national distributors. Suddenly people couldn鈥檛 rely on interpersonal relationships or cultural norms to safeguard their transactions; they didn鈥檛 know, and often never even met, the 颅people they were doing business with. The result, UCLA sociologist Lynne Zucker has argued, was the destruction of the trust that had sustained the US economy up until that point.

In the ensuing years, formal systems sprang up as proxies for the trust that citizens had lost in one another. The decades between 1870 and 1920 saw the explosion of the 鈥渟ocial overhead capital sector鈥濃攊ndustries like banking, insurance, and legal services that established rules and backstops for the new business environment. Meanwhile, a slate of government regulations helped establish the rules that this new breed of corporations had to follow. 鈥淭hrough institutionalizing socially created mechanisms for producing trust,鈥 Zucker writes, 鈥渢he economic order was gradu颅ally reconstructed.鈥 The casual, intimate, interpersonal form of trust was replaced by a centralized system of codified safeguards.

But the problem with institutionalized trust is that it can be, in tech industry parlance, a high-friction affair. eBay couldn鈥檛 require everyone with a few extra Beanie Babies to go through the regulatory rigmarole of establishing themselves as a licensed shopkeeper. So over several years, Chesnut鈥檚 team built its own trust infrastructure. It began monitoring the activity across the eBay marketplace, flagging potentially problematic sellers or buyers, providing its own payment options, and eventually guaranteeing every purchase. In so doing, eBay evolved from a passive host to an active participant in every transaction. Like the explosion of institutional banking and insurance in the early 20th century, this new system acted as a trust proxy; it didn鈥檛 require people to trust one another, because they could rely on a centralized system to protect their interests.

That process has been recapitulated at companies like Airbnb. Initially, cofounders Brian Chesky, Joe Gebbia, and Nate Blecharczyk imagined the service as a kind of event-颅specific craigslist, pairing renters with hosts and then leaving them to their own devices. But over the years, the company broadened its scope and took on a larger and larger role鈥攈andling all of the payments, hosting reviews, hiring professional photographers to shoot properties, and providing a platform for hosts and guests to communicate with one another. The biggest ramp-up came after the infamous 鈥渞ansackgate鈥 incident of June 2011, in which a host named EJ found her San Francisco apartment trashed by guests who stole her jewelry, hard drive, passport, and credit cards. In response, Airbnb instituted many new security provisions, set up a 24/7 customer-service hotline, established a $50,000 host guarantee鈥攍ater increased to $1 million鈥攁nd built a new trust and safety division.

Anna Steel, a former government investigator who now serves as one of Airbnb鈥檚 lead trust and safety 颅managers, was hired nearly a year after the crisis. Today she heads a team of 15 case managers, part of an 80-person group with offices in Singapore, Dublin, and San Francisco. To get a sense of her work, I drop in on a planning meeting in advance of SXSW, the Austin music and technology festival that has become one of the company鈥檚 most high-颅volume events. Airbnb鈥檚 conference rooms are famous for their elaborate decor鈥攅ach a re-颅creation of an actual Airbnb property鈥攂ut this meeting is held in a drab, unadorned space. One by one, the members of a four-woman task force discuss their progress. Emily Gonzales has been reaching out to guests that the company鈥檚 system has flagged as posing the greatest property-damage risk鈥攍arge groups or first-time renters who have booked rooms in swanky homes鈥攖o remind them to take care of their hosts鈥 property. Jaspreet Bansal, who left her job as a criminal prosecutor in Newark in December, has been working with agents to scan the site for potentially illegitimate listings. Meanwhile, Brittany Galvan is planning to head out to Austin to handle any problems that arise.

They are aided in these efforts by the huge pile of data the company has amassed. Every element of a booking鈥攖he reservation, payment, communication between host and guest, and review鈥攖akes place through Airbnb鈥檚 platform so the company can track each stay from conception to completion. If a host uses the words Western Union in a conversation with a guest鈥攁 sign that they may be trying to route around Air颅bnb鈥檚 system鈥攖he company will block the message. If a host and guest are repeatedly booking rooms with one another, it could be a scam to build up fake positive reviews. And if a new host pops up and instantly starts booking expensive reservations with a new user, that could signal something like a money-颅laundering racket. Airbnb鈥檚 analytics system takes factors like these into account, then assigns each reservation a 鈥渢rust score.鈥 If the score is too low, it鈥檚 automatically flagged for further investigation. (The system isn鈥檛 foolproof. In March a comedian discovered that his house had been used for a massive sex party. But Airbnb says it is largely successful; of 6聽million guests in 2013, the company paid out only 700 host claims.) In a lot of ways, this process is simi颅lar to the trust infrastructure that eBay developed鈥攁 machine that assumes risk on behalf of its customers and frees them from the responsibility of assessing each other鈥檚 trustworthiness.

But here鈥檚 the thing: eBay is a pretty binary experience. You either get what you ordered or you don鈥檛. For a system like that, this kind of centralized trust infrastructure is sufficient. It helps weed out fraudsters and incompetents. Similarly, licensing departments and health inspectors help to guarantee a baseline level of safety and security. You can check into a licensed hotel knowing you are in fact entering a hotel and not an organ-颅harvesting lab that looks like a hotel. But they can鈥檛 guarantee you鈥檒l have a good experience鈥攖hat the bellhop won鈥檛 be a jerk or room service won鈥檛 bring you a lukewarm omelet. That鈥檚 up to the hotel company, which manages its staff to provide a standard of service.

But sharing-economy companies don鈥檛 have on-site managers and staffs. They鈥檙e a ragtag collection of loosely organized individuals. Their centralized trust infrastructures may catch obvious bad actors鈥攑urveyors of fake listings, money launderers, thieves鈥攂ut they won鈥檛 stop more run-of-the-mill offenders like the driver who鈥檚 got a bit of a lead foot or the houseguest who carelessly drips candle wax all over your speaker. That requires more subtle forms of social engineering. RelayRides CEO Andre Haddad compares it to parenthood. 鈥淚 have three kids,鈥 he says. 鈥淵ou can鈥檛 control them, but you want to nudge them to do the right thing.鈥

And like parents, many companies are making up the rules as they go鈥攁nd sometimes learning new tricks by accident. When Haddad joined RelayRides in September 2011, the company was pursuing a 颅Zipcar-like model. Customers rented cars by the hour and never met the owners. They accessed and started their rentals by swiping a membership card past a reader that the company had installed in every owner鈥檚 car. But by spring 2012, RelayRides needed to make some changes. For one thing, it was clear that Avis and Hertz had a more appealing model than Zipcar; the market for traditional car rentals is nearly 60 times larger than the hourly car-sharing rental market. And the company wanted to grow around the globe, making it impractical and expensive to set up a complicated hardware installation for every new member. So, as of March 2012, RelayRides ditched the card reader. Instead, renters and owners began meeting in person to hand off keys and look over the 颅vehicle.

The results, Haddad says, were striking. RelayRides was just looking for a more convenient, cost-effective way to expand its business. But it turned out that the face-to-face meeting caused renters to take better care of the cars鈥攁nd it made the experience better for both parties. Owners made significantly fewer damage claims under the new approach, and both renters and owners reported much higher satisfaction rates after meeting in person. 鈥淭hey really liked that human connection,鈥 Haddad says. 鈥淧eople strike up a conversation and realize they have something in common, which boosts trust and makes people feel accountable. They鈥檙e going to have to return this car to that person and look them in the eye.鈥

Ultimately, this is what separates companies like RelayRides from the eBay-like person-to-person marketplaces that came before. When you buy a camera on eBay, you only know your seller as NikonIcon1972. In the sharing economy, we aren鈥檛 anonymous. We may not meet our trading partners face-to-face, as in the RelayRides example. But because our transactions are often linked through our Facebook accounts鈥攕ome version of our real identities鈥攚e are dealing, even virtually, with real people. It鈥檚 a digital re-颅creation of the neighborly interactions that defined pre-颅industrial society. Except that now our neighbor is anyone with a Facebook account.

Most of these marketplaces try to maximize that feeling of interpersonal connection. That鈥檚 why Lyft鈥攕logan: 鈥淵our friend with a car鈥濃攅ncourages riders to sit in the front seat like a friend rather than in the backseat like a fare. It鈥檚 why Airbnb hosts are asked to include large photos of themselves on their profiles, and why the company urges hosts and guests to communicate with each other before every stay. It鈥檚 why the Feastly website includes personal biographies of every chef and encourages pre-dinner-party communication. 鈥淭here are psychological studies up the wazoo about how we mistrust people when we don鈥檛 know them,鈥 says Charles Green, a trust expert who advises companies like Shell and Accenture. 鈥淏ut we don鈥檛 mess with people we know.

Introducing people to one another may encourage them to behave better鈥攊t may reduce insurance payouts and help a company鈥檚 bottom line. But it also makes for a radically different experience than we鈥檝e come to expect from our service economy. In my conversations with Lyft riders and drivers, practically everyone said some version of the following: 鈥淚 like dealing with real 颅people.鈥 Of course, the licensed cabbie is a real person. So is the bellhop, the line cook, the kennel owner. But when we interact with them, they are operating as agents of a commercial enterprise. In the sharing economy, the commerce feels almost secondary, an afterthought to the human connection that undergirds the entire experience. (This is due in part to the fact that the payment itself so often happens electronically and invisibly.) In this way, it suggests a return to pre-industrial society, when our relationships and identities鈥攕ocial capital, to use the lingo鈥攎attered just as much as the financial capital we had to spend.

That鈥檚 the carrot side of a more intimate economy, the idea that treating people well will result in a better experience. There is a stick side as well: Act badly and you鈥檒l be barred from participat颅ing. Nick Grossman, a general manager at Union Square Ventures and a visiting scholar at the MIT Media Lab, says that while Uber drivers are generally positive about the service, he has spoken with some who worry about picking up a 颅couple of bad reviews, falling below the acceptable rating threshold, and getting fired. (The same holds for passengers: Manit, the Lyft driver, says she won鈥檛 pick up anyone with less than a 4.3-star rating.) 鈥淭here鈥檚 a legitimate question: How do we feel about living in an environment of hyper-accountability?鈥 Grossman asks. 鈥淚t鈥檚 very effective at producing certain outcomes. It鈥檚 also very Darwinian.鈥 Just like resi颅dents of pre-industrial America, sharing-economy participants know that every transaction contributes to a reputation that will follow them, potentially for the rest of their lives.

Indeed, for the time being the boundaries of the sharing economy are protected fairly rigidly. If you鈥檝e ever been caught driving more than 20 miles over the speed limit, you can鈥檛 rent a car on RelayRides. Aspiring Lyft drivers must pass a background and DMV check and get approved by a mentor, who judges applicants not just on driving ability but on personality. DogVacay hosts go through a five-step vetting process that includes training videos, quizzes, and a telephone interview.

More broadly, new sharing economy companies are most likely to draw from a set of like-minded, forward-thinking early adopters. That dynamic undoubtedly has helped hosts and drivers trust their customers; studies show that we are more liable to trust 颅people who seem to share our values and personal traits. (鈥淚 don鈥檛 know if I鈥檇 do this in Philly,鈥 a San Francisco Lyft driver named Joel confesses. 鈥淏ut here everybody鈥檚 so nice.鈥) It could also explain a troubling study from two Harvard Business School professors showing that Airbnb guests pay black hosts less than their white counterparts. (The authors鈥 suggested solution: de-颅emphasize hosts鈥 profile photos, which flies in the face of the company鈥檚 trust-building efforts.)

But in the end, these new mechanisms for creating and safeguarding interpersonal trust may have the power to make us comfortable with people and experiences we never would have otherwise considered. Kari Sweetland, a 30-year-old HR coordinator, recently signed up for Tinder, the wildly popular hook-up app. Tinder isn鈥檛 normally considered part of the sharing economy, but it does operate by some of the same logic. Users meet potential paramours, not by answering lengthy questionnaires but by simply linking to their Facebook accounts and swiping through a series of photos. Tinder鈥檚 algorithm displays people nearby, noting who shares interests or social connections and, if both parties approve one another, lets them send messages through the app until they feel comfortable enough to meet in person.

When Sweetland first signed up for Tinder, she says, she had a moment of hesitation, the vague sense that what she was about to do was a little crazy, meeting strangers based on nothing more than the swipe of a touchscreen. But then she remembered that she did something similar every time she stepped into a Lyft car or stayed at an Airbnb鈥攚hich she and her friends do all the time. Suddenly, meeting a stranger didn鈥檛 seem like such a scary risk after all. It was just a regular way for her to interact with her fellow San Franciscans. 鈥淚鈥檝e accepted that in my life, and everyone here has too,鈥 she says. This isn鈥檛 oddball behavior, the actions of someone flaunting social norms. It鈥檚 just what people do. 颅

Four score and 50,000 years ago, the first chump was scammed. Since then, we鈥檝e developed norms, structures, and safeguards to protect us while trading, even with strangers. 鈥擩ulia Greenberg

50,000 BC

Friend-to-Friend Buyers and sellers mainly barter and trust only friends. Trading is based on reciprocity and reputation; cheaters are shunned.

8000 BC

Neighbor-to-Neighbor Small villages form; 颅people begin trading with neighbors. It鈥檚 easier to trust someone when you know where they live.

1200 BC

Currency Portable currency聽develops鈥攆irst as shells, then as coins鈥攁nd becomes a shared medium of exchange, replacing barter.

AD 650

Paper Money Cash with no inherent value is popularized for the first time. 颅People trust that bits of paper can be exchanged for real goods.

AD 1000

Stranger-to-Stranger Buyers and sellers begin to trade with strangers through trusted intermediaries who bring goods to market.


Person-to-颅company As corporations replace local stores and markets, trust comes from regulations, insurers, banks, and law firms.


Person-to-world 颅People buy goods from multi颅national corpora颅tions. Watchdog groups, along with global regulations and banks, secure these interactions.


Networked Products are purchased through websites built by companies that provide a centralized marketplace and trust infrastructure.


Intimate 颅New mechanisms emerge to secure in-颅person transactions that are brokered through digital marketplaces.