杏吧原创

THE first thing bought on eBay was a pair of ice skates. They came with a handwritten note: 鈥淚 hope your daughter enjoys these as much as my daughter did.鈥� It made Zak鈥檚 day.

As someone who studies the neurological basis of trust, Zak knew exactly what was going on. Feel-good chemicals had flooded his bloodstream, changing how he felt about a stranger over the internet. But that didn鈥檛 stop him leaving a shining review.

Human interactions are built on trust. We trust others to hand over the goods when we pay them. We trust banks with our money and doctors with our lives. We trust governments to run our countries and newspapers to tell us how they are doing it. The more trust in a society, the better it fares. Put another way: without trust, society would collapse.

But something strange is happening. Public trust in our institutions has plummeted in the past decade. Nearly half of people in the US , according to a poll carried out in June. In the UK, trust the press. And yet we are putting more trust than ever before in people we meet on the internet. The sharing economy is booming. It is normal to invite strangers to sleep on our sofas, meet us for dates, pick us up in their cars and look after our pets. The internet has brought us to a tipping point, fundamentally changing who we trust and why. Technology allows us to make informed decisions and vet individuals. But should we really be letting our guard down?

Trust is a human instinct that is essential to our survival. It first evolved when we lived in small tribal groups, and probably provided benefits in times of conflict. Groups that were better at working together 鈥� more trusting 鈥� were more likely to survive than less cooperative rivals.

鈥淎ll societies are based on trust because you can鈥檛 do everything yourself,鈥� says , who studies online trust at the Oxford Internet Institute in the UK. 鈥淎t some point you have to trust someone to keep the fire going.鈥�

The trouble is, trusting groups can always be exploited by untrustworthy individuals. Putting your trust in someone puts you at risk, it makes you vulnerable. 鈥淭rust is the bridge between the known and the unknown,鈥� says at the University of Oxford鈥檚 Said Business School.

Given the ever-present possibility of betrayal, how do we manage to cross the bridge? The hormone oxytocin plays a critical role. Known for being an all-round social fixer, it kicks in to make us feel good when we put our faith in others.

Zak has been studying the effect of oxytocin on trust for 15 years. In a series of experiments, his team at Claremont Graduate University in California measured blood levels of the hormone in people playing a game in which they had to cooperate with strangers to win money. Each player was given a sum of money and could choose either to keep it or give some or all of it to a second player. They, in turn, could choose either to reward the first player鈥檚 generosity by returning a portion of the cash or keep it all for themselves.

鈥淎ll societies are based on trust 鈥� you can鈥檛 do everything yourself鈥�

The traditional view in behavioural economics is that only a sucker trusts others because your trust won鈥檛 be reciprocated. But Zak鈥檚 blew a hole in that idea. In the initial study of 38 people, who knew that they were being trusted by the other player, participants shared more than half of their money, on average. Indeed, most people responded in kind when others trusted them 鈥� an effect that was directly tied to a surge in oxytocin. Almost everyone 鈥� 95 per cent of people in those first studies 鈥� showed a bump in oxytocin when others trusted them. Zak has now run the tests in hundreds of people. 鈥淚t鈥檚 a near-universal response,鈥� he says.

Later studies confirmed the influence of the hormone: when players were given synthetic oxytocin in a nasal spray, they handed over more cash than those given a placebo.

Our oxytocin response to cues such as body language and smells takes about a second to kick in. This lets us make snap judgements about who is and isn鈥檛 trustworthy. 鈥淲e鈥檙e hyper-aware of the social environment we鈥檙e in,鈥� says Zak. 鈥淲hen we meet someone new we just get a sense 鈥� 鈥極h, this guy鈥檚 great鈥� or 鈥業 don鈥檛 know, I鈥檝e got a bad feeling about this person鈥�.鈥�

All of this is fine for a society based on face-to-face interaction, as ours was for tens of thousands of years. Then came the industrial revolution. 鈥淲e realised that trust didn鈥檛 scale to international trade,鈥� says Botsman. So we invented proxies. Trust started to flow through intermediaries like insurers and lawyers. The local manager became the face of the big bank; our trust in them extended to the organisation. Just as we had put more trust in friends of friends, reputation came to stand in for direct knowledge in fostering our trust in institutions. Eventually, marketing teams became adept at creating brand personalities that encourage us to anthropomorphise faceless corporations.

And more recently, companies have been fronted by celebrity founders like Mark Zuckerberg or Steve Jobs. Zuckerberg often addresses his billion-plus users in video messages, for example. When Facebook was criticised for dismissing claims that it played a role in the 2016 US presidential election, it was Zuckerberg who issued a personal apology.

In the past, it was easier to put our faith in institutions because their failings were hidden from view. Then came the internet. Global scandals like the 2008 banking crisis, the US National Security Agency鈥檚 over-reaching surveillance programme and Volkswagen鈥檚 emissions-test cheats spread around the world in hours and were dissected by millions online. Of course distrust in big institutions predates the internet, but technology has made it an international sport. It is easier than ever before for leaks to become common knowledge.

And there are many more sources now. Opinion is no longer shaped only by journalists, experts or state authorities. With constant access to a deluge of information, rather than putting our trust in the institutions our peers also trust, as we once did, we鈥檙e now trusting our peers instead of those institutions.

With the curtains pulled back and gossip in full flow, people start to question the entire system. 鈥淭he cycle of questioning and doubt puts trust in a precarious place,鈥� says Botsman. This erosion of trust becomes contagious and spreads quickly. 鈥淔ear and suspicion are its enemies,鈥� she says.

So how is it, in this climate, that companies which rely on people placing their faith in strangers are thriving? The trick seems to be forging links directly between individuals, while appearing to cut out the institutions.

When many of these companies started out, they could still rely partly on existing social connections. You might stay at an Airbnb based on a personal recommendation, for example. But as they expanded, it became more and more crucial to encourage trust between users.

It was eBay that made people realise you could trust someone online. 鈥淚 remember talking to colleagues who said it was never going to work,鈥� says Floridi. Many assumed trust required more than a few typed words of assurance from a stranger. 鈥淚t鈥檚 like taking away the handshaking and keeping only the verbal agreement,鈥� he says. What the internet offers instead is information.

Companies like , and 鈥� which put you in contact with strangers you may choose to invite into your home 鈥� have managed to foster trust in a new way. 鈥淥ur objective is to provide a platform that lets people provide each other with information that will strengthen their relationship,鈥� says , CEO of TaskRabbit. Information comes in the form of profiles, pictures, personal details.

鈥淣ormally, I wouldn鈥檛 want a stranger knowing where I live,鈥� says Zak. 鈥淏ut we have so much information about that person and they have so much information about us we can both make pretty informed decisions.鈥�

In most situations we are more likely to trust a stranger if we have something in common, such as age, gender or background. But online ratings override that tendency. In a 2016 study, and colleagues at Stanford University in California interactions between Airbnb users, consisting of requests to stay and acceptances by hosts. They found that hosts with better reputations 鈥� measured by the number of positive reviews 鈥� were contacted by a more diverse group of people than hosts with less-established reputations, who were more likely to be contacted by people from a similar demographic. So the reputation system can help overcome ingrained biases. A similar pattern was found for hosts accepting guests.

鈥淭rust is going down in most places but we鈥檙e seeing people trust each other more,鈥� says Nick Shapiro, the global director of trust and risk management at Airbnb.

In fact, many of these companies have come to realise that trust itself is their product. That is why Airbnb created a department dedicated to trust and risk management 鈥� and brought in Shapiro, who has worked for the CIA and has a background in security, to head it up.

Hawking trust between individuals requires some sleight of hand: we are more likely to trust people at a distance when they are backed up by trustworthy organisations. We trust strangers on Airbnb far more than strangers on a marketplace such as Craigslist, for instance.

The safety net

The organisations most adept at fostering trust in strangers do so in various ways 鈥� by providing guarantees or insurance policies if things go wrong, or policing who we connect with to begin with. For instance, Airbnb checks all hosts and guests aren鈥檛 on terrorist, sanctions or regulatory watch lists, and also runs background screening for things like prior felony convictions or sex offender registrations. For every reservation, they run an automated risk assessment that evaluates hundreds of signals to flag suspicious activity before the people they have introduced ever actually meet. 鈥淚t鈥檚 constantly being tweaked,鈥� says Shapiro. 鈥淚t gets smarter every single day.鈥�

Airbnb can also tell if you are who you say you are. After a spate of account takeovers, where usernames and passwords stolen from other sites were used to log into accounts, the company acquired a start-up called Trooly to avoid scams. Trooly鈥檚 software digs through the digital trail we all leave behind 鈥� from social media to police records to the dark web 鈥� to determine how trustworthy we are. The company even claims its software can predict how an individual is likely to behave. Is this person going to trash your house? 鈥淚t might completely freak us out, but it does help weed out who is not trustworthy,鈥� says Botsman. 鈥淚t takes us full circle back to one-to-one connections on a scale that was never possible before.鈥�

By filtering out bad actors, the system provides users with a pool of vetted individuals. It resembles the way we are more likely to trust friends of a friend 鈥� they are vouched for.

Where does that leave us? 鈥淚n a few years鈥� time, people will just call what we do the normal way to get things done,鈥� says Brown-Philpot. It may be that these systems help to ratchet up trust levels across society by lowering our guard. We are being conditioned to get into a stranger鈥檚 car or post ice skates to a person who has promised to pay you, says Zak.

A major concern is that we will become overly dependent on digital platforms to manage trust for us. 鈥淧eople trust people, not institutions,鈥� recently said. That is misleading. We trust people online because of the institutions 鈥� Facebook included 鈥� that make it possible.

If we continue to outsource trust in this way, we must also ask: how reliable are these new institutions? For instance, London鈥檚 transport authority has not renewed Uber鈥檚 operating licence in part because the company allegedly failed to report a series of sexual assaults to the police. The accusation directly threatens to shake Uber鈥檚 network of trust.

鈥淲e may end up with a shallower kind of trust 鈥� not making the leap of faith鈥�

By allowing internet companies to determine the trustworthiness of others, we no longer need to rely on our own instincts. 鈥淲ith more information we don鈥檛 need to trust with blind faith,鈥� says Floridi. 鈥淲e can take people at face value because we can check up on them.鈥�

That鈥檚 no bad thing. But we may end up with a shallower kind of trust: if something is completely transparent, you never have to make the leap of faith. 鈥淵ou鈥檝e actually given up on trust,鈥� says Botsman. When online companies ramp up their security and assurances to protect their users 鈥� and themselves 鈥� from harm, we no longer need to build mutual trust, says Coye Cheshire at the University of Berkeley, California. He argues that we are left with a paradox: the systems put in place to foster trust between strangers online may undermine the need for trust in the first place.

Of course, the truth is, like our instincts, these systems are not foolproof. Bad apples can sneak by the algorithms just as they dupe us. So we need to take care not to simply offload decisions about who or what to trust on to algorithms (see 鈥�In bots we trust鈥�), and subject the companies that write them to the same level of scrutiny as the institutions we have so recently come to mistrust. Democratic governments may disappoint and betray us, but we can hold them to account at the voting booth. Private companies trading in trust must be held to account as well.

鈥淥ne of the things we need to realise now is that efficiency and speed are in fact the enemy of trust,鈥� says Botsman. 鈥淏ut that鈥檚 really difficult to do in an age when we want a car in two minutes and a date by lunch time.鈥�

This article appeared in print under the headline 鈥淚n what? we trust鈥�