by Jon Ronson
‘Are you sure you want to say that Walmart was soul-sucking?’ Farukh said.
‘Oh … What? Really?’ Lindsey laughed as if to say, ‘Come on! Everyone knows that about Walmart!’ But then she hesitated.
The conference call was proving an unexpectedly melancholic experience. It was nothing to do with Farukh. He really felt for Lindsey and wanted to do a good job for her. The sad thing was that Lindsey had incurred the Internet’s wrath because she was impudent and playful and foolhardy and outspoken. And now here she was, working with Farukh to reduce herself to safe banalities - to cats and ice cream and Top 40 chart music. We were creating a world where the smartest way to survive is to be bland.
*
There was a time when Michael Fertik wouldn’t have needed to be so calculating. Back in the mid 1990s search engines were only interested in how many times a particular keyword appeared within a page. To be the number-one Jon Ronson search term on AltaVista or HotBot you just had to write Jon Ronson over and over again. Which for me would be the most fantastic website to chance upon, but for everyone else, less so.
But then two students at Stanford, Larry Page and Sergey Brin, had their idea. Why not build a search engine that ranked websites by popularity instead? If someone is linking to your page, that’s one vote. A link, they figured, is like a citation - a nod of respect. If the page linking to your page has a lot of links into it, then that page counts for more votes. An esteemed person bestowing their admiration upon you is worth more than some loner doing the same. And that was it. They called their invention PageRank, after Larry Page, and as soon as they turned the algorithm on, us early searchers were spellbound.
This was why Farukh needed to create LinkedIn and Tumblr and Twitter pages for Lindsey. They come with a built-in high PageRank. The Google algorithm prejudges them as well liked. But for Michael the problem with Google is that it is forever evolving - adjusting its algorithm in ways it keeps secret.
‘Google is a tricky beast and a moving target,’ Michael told me. ‘And so we try to decipher it, to reverse-engineer it.’
This was what Michael knew right now: ‘Google tends to like stuff that’s old. It seems to think old stuff has a certain authority. And Google tends to like stuff that’s new. With the intervening stuff, week six, week twelve, there’s a dip.’ Which was why Michael’s people predicted that Lindsey’s love of cats or whatever would achieve ‘initial strong impact’, followed by ‘fluctuation’. And after fluctuation: ‘reversion’.
Michael’s clients dread reversion. There’s nothing more dispiriting than seeing the nice new judgements disappear down to page two and the horrific old judgements bubble back up again. But reversion is actually their friend, Jered Higgins told me. Reversion is when you think Glenn Close is dead but she suddenly leaps up in the bath, apparently filled with a new violent fervour, but really she’s muddled and wounded and vulnerable.
‘Reversion shows that the algorithm is uncertain,’ Jered said. ‘It’s the algorithm shifting things around and wondering what, from a mathematical standpoint, is the story that needs to be told about this person.’
And during this uncertainty, Jered said, ‘We go in and blast it.’
The blasting - the bombardment of the algorithm with Tumblr pages about Lindsey’s trips to the beach, the Shock and Awe of these pleasant banalities - has to be choreographed just right. Google knows if it’s being manipulated. Alarm bells go off. ‘So we have a strategic schedule for content creation and publication,’ Jered said. ‘We create a natural-looking activity online. That’s a lot of accumulated intelligence.’
*
Michael Fertik took me for dinner and talked to me about the criticism people often level at him, that ‘any change of search results is manipulating truth and chilling free speech’. He drank some wine. ‘But there is a chilling of behaviour that goes along with a virtual lynching. There is a life modification.’
‘I know,’ I said. ‘For a year Lindsey Stone had felt too plagued to even go to karaoke.’ And karaoke is something you do alone in a room with your friends.
‘And that’s not an unusual reaction,’ Michael said. ‘People change their phone numbers. They don’t leave the house. They go into therapy. They have signs of PTSD. It’s like the Stasi. We’re creating a culture where people feel constantly surveilled, where people are afraid to be themselves.’
‘Like the NSA,’ I said.
‘This is more frightening than the NSA,’ said Michael. ‘The NSA is looking for terrorists. They’re not getting psychosexual pleasure out of their schadenfreude about you.’
I wondered what to make of Michael’s Stasi analogy. There’s an old Internet adage that as soon as you compare something to the Nazis you lose the argument. Maybe the same could be said about the Stasi - the East Germans’ secret police force during the Cold War. They would, after all, creep into the homes of suspected enemies of the state and spray radiation onto them as they slept, their idea being to use the radiation as a tracking device. Stasi agents would follow them through crowds, pointing Geiger counters at them. A lot of suspected enemies of the state died of unusual cancers during the Stasi’s reign.
But the Stasi weren’t just about inflicting physical horror. Their main endeavour was to create the most elaborate surveillance network in world history. It didn’t seem unreasonable to scrutinize this aspect of them in the hope it might teach us something about our own social media surveillance network.
In Anna Funder’s seminal history of the Stasi - Stasiland - she interviews a woman named Julia who was one day called in for interrogation. The Stasi had intercepted love letters between her and her Western boyfriend. They were sitting on the officer’s desk in the interrogation room.
There was a pile of her letters to the Italian. There was a pile of his letters back to her. This man knew everything. He could see when she had doubts. He could see by what sweet-talking she had let herself be placated. He could see the Italian boyfriend’s longing laid bare.
-Anna Funder, Stasiland, Granta, 2003
Julia told Anna Funder that she was ‘definitely psychologically damaged’ by the incident - the way the officer read through her letters in front of her, making little comments. ‘That’s probably why I react so extremely to approaches from men. I experience them as another possible invasion of my intimate sphere.’
Anna Funder wrote Stasiland back in 2003 - fourteen years after the fall of the Stasi and three years before the invention of Twitter. Of course no prurient or censorious bureaucrat had intercepted Justine Sacco’s private thoughts. Justine had tweeted them herself, labouring under the misapprehension - the same one I laboured under for a while - that Twitter was a safe place to tell the truth about yourself to strangers. That truth-telling had really proven to be an idealistic experiment gone wrong.
Anna Funder visited a Stasi officer whose job had been to co-opt informants. She wanted to know how - given that informant pay was terrible, and the workload was ever burgeoning, with more and more behaviours being redefined as enemy activities - he manage to persuade people to get on board.
‘Mostly people just said yes,’ he told her.
‘Why?’ she asked him.
‘Some of them were convinced of the cause,’ he said. ‘But I think mainly because informers felt they were somebody, you know? Someone was listening to them for a couple of hours every week, taking notes. They felt they had it over other people.’
That struck me as a condescending thing for the Stasi man to say about his informants. And it would be a condescending thing to say about Twitter users too. Social media gives a voice to voiceless people - its egalitarianism is its greatest quality. But I was struck by a report Anna Funder discovered that had been written by a Stasi psychologist tasked with trying to understand why they were attracting so many willing informants. His conclusion: ‘It was an impulse to make sure your neighbour was doing the right thing.’
*
In October 2014 I took a final drive
up to visit Lindsey Stone. Four months had passed since I’d last spoken to her or Farukh - I hadn’t called them and they hadn’t called me - and given that they’d only taken her on for my benefit I’d half-wondered if maybe it had all been quietly wound down in my absence.
‘Oh God no,’ said Lindsey. We sat at her kitchen table. ‘They call me every week, week after week. You didn’t know that?’
‘No,’ I said.
‘I thought you guys were talking all the time,’ she said.
Lindsey got out her phone and scrolled through her innumerable emails from Farukh. She read out loud some blogs his team had written in her voice, about how it’s important when traveling to use the hotel safe - ‘Stay alert, travelers!’ - and how if you’re in Spain you should try the tapas.
Lindsey got to pre-approve everything and she’d only told them no twice, she said, to the blog about how much she’s looking forward to Lady Gaga’s upcoming jazz album (‘I like Lady Gaga, but I’m not really excited about her jazz album’) and to her tribute to Disneyland on the occasion of its fiftieth birthday: ‘Happy Birthday Disneyland! The Happiest Place on Earth!’
‘Happy Birthday Disneyland!’ Lindsey blushed. ‘I would never … I mean, I had a great time at Disneyland …’
‘Who doesn’t?’ I said.
‘But still …’ Lindsey trailed off.
After we both laughed about the Happy Birthday Disneyland blog we both stopped laughing and felt bad.
‘They’re working so hard,’ Lindsey said.
‘And it’s what they have to do,’ I said.
‘Yeah,’ Lindsey said. ‘One of my friends from high school said, “I hope it’s still you. I want people to know how funny you are.” But it’s scary. After all that’s happened, what’s funny to me … I don’t want to go anywhere near the line, let alone cross it. So I’m constantly saying, ‘I don’t know, Farukh, what do you think?’
‘This journey started with my identity being hijacked by a spambot,’ I said. ‘Your personality has been taken by strangers twice now. But at least this second time around it’s nice.’
Lindsey hadn’t typed her name into Google for eleven months. The last time had been a shock. It was Veterans’ Day and she discovered some ex-army people ‘wondering where I was and not in a good way’.
‘They were thinking about tracking you down so they could re-destroy you?’ I asked.
‘Yeah,’ she said.
She hadn’t looked since. And now she swallowed and began to type: L … I … N …
Lindsey shook her head, stunned. ‘This is monumental,’ she said.
Two years ago the photograph stretched to Google Images horizon - uninterrupted, mass-production shaming, ‘pages and pages and pages,’ Lindsey said, ‘repeating endlessly. It felt so huge. So oppressive.’
And now: gone.
Well - nearly gone. There was still a scattering of them, maybe three or four, but they were interspersed with lots of photographs of Lindsey doing nothing bad. Just smiling. Even better, there were lots of photographs of other Lindsey Stones - people who weren’t her at all. There was a Lindsey Stone volleyball player, a Lindsay Stone competitive swimmer. The swimmer had been captured mid-stroke, moments from winning the New York State 500-yard freestyle championship. It was captioned, ‘Lindsay Stone had the right plan in place and everything was going exactly to plan.’
A whole other person, doing something everyone could agree was lovely and commendable. There was no better result than that.
15
YOUR SPEED
We have always had some influence over the justice system, but for the first time in 180 years - since the stocks and the pillory were outlawed - we have the power to determine the severity of some punishments. And so we have to think about what level of mercilessness we feel comfortable with. I, personally, no longer take part in the ecstatic public condemnation of anybody, unless they’ve committed a transgression that has an actual victim, and even then not as much as I probably should. I miss the fun a little. But it feels like when I became a vegetarian. I missed the steak, although not as much as I’d anticipated, but I could no longer ignore the slaughterhouse.
I kept remembering something Michael Fertik had said to me at the Village Pub in Woodside. ‘The biggest lie,’ he said, ‘is “The Internet is about you.” We like to think of ourselves as people who have choice and taste and personalized content. But the Internet isn’t about us. It’s about the companies that dominate the data flows of the Internet.’
Now, I suddenly wondered. Did Google make money from the destruction of Justine Sacco? Could a figure be calculated? And so I joined forces with a number-crunching researcher, Solvej Krause, and began writing to economists and analysts and online ad revenue people.
Some things were known. In December 2013, the month of Justine’s annihilation, 12.2 billion Google searches took place - a figure that made me feel less worried about the possibility that people were sitting inside Google headquarters personally judging me. Google’s ad revenue for that month was $4.69 billion. Which meant they made an average of $0.38 for every search query. Every time we typed anything into Google: 38 cents to Google. Of those 12.2 billion searches that December, 1.2 million were people searching the name Justine Sacco. And so, if you average it out, Justine’s catastrophe instantaneously made Google $456,000.
But it wouldn’t be accurate simply to multiply 1.2 million by $0.38. Some searches are worth far more to Google than others. Advertisers bid on ‘high yield’ search terms, like ‘Coldplay’ and ‘Jewellery’ and ‘Kenya vacations’. It’s quite possible that no advertiser ever linked their product to Justine’s name. But that wouldn’t mean Google made no money from her. Justine was the worldwide number-one trending topic on Twitter. Her story engrossed social media users more than any other that night. I think people who wouldn’t otherwise have gone onto Google did so specifically to hunt for her. She drew people in. And once they were there I’m sure at least a few of them decided to book a Kenya vacation or download a Coldplay album.
I got an email from the economics researcher Jonathan Hersh. He’d come recommended by the people who make Freakonomics Radio on WNYC. Jonathan’s email said the same thing: ‘Something about this story resonated with them, so much so that they felt compelled to google her name. That means they’re engaged. If interest in Justine were sufficient to encourage users to stay online for more time than they would otherwise, this would have directly resulted in Google making more advertising revenue. Google has the informal corporate motto of “don’t be evil”, but they make money when anything happens online, even the bad stuff.’
In the absence of any better data from Google, he wrote, he could only ever offer a ‘back of the envelope’ calculation. But he thought it would be appropriately conservative - maybe a little too conservative - to estimate Justine’s worth, being a ‘low-value query’, at a quarter of the average. Which, if true, means Google made $120,000 from the destruction of Justine Sacco.
Maybe that’s an accurate figure. Maybe Google made more, or possibly less. But one thing’s certain. Those of us who did the actual annihilating? We got nothing.
*
From the beginning I’d been trying to understand why - once you discount Gustave Le Bon and Philip Zimbardo’s theories of viruses and contagion and evil - online shaming is so pitiless. And now I think I have the answer. I found it in, of all places, an article about a radical traffic-calming scheme tested in California in the early 2000s. The story - by the journalist Thomas Goetz - is a fantastically esoteric one. Goetz writes about how in the school zones of Garden Grove, California, cars were ignoring speed signs and hitting ‘bicyclists and pedestrians with depressing regularity’. And so they tried something experimental. They tried Your Speed signs.
After I read Thomas Goetz’s article about Your Speed signs I spent a long time trying to track down their inventor. He turned out to be an Oregon road-sign manufacturer called Scott Kelley.
‘I
remember exactly where I was when I thought of them,’ he told me over the telephone. ‘It was the mid 1990s. I was over by my girlfriend’s house. I was driving through a school zone. And my mind just pictured one of the signs up on a pole.’
‘What made you think they’d work?’ I asked him. ‘There was nothing about them to suggest they’d work.’
‘Right,’ said Scott. ‘And that’s where it gets interesting.’
They really, logically, shouldn’t have worked. As Thomas Goetz wrote,
The signs were curious in a few ways. They didn’t tell drivers anything they didn’t already know - there is, after all, a speedometer in every car. If a motorist wanted to know their speed, a glance at the dashboard would do it … And the Your Speed signs came with no punitive follow-up - no police officer standing by ready to write a ticket. This defied decades of law-enforcement dogma, which held that most people obey speed limits only if they face some clear negative consequence for exceeding them.
In other words, officials in Garden Grove were betting that giving speeders redundant information with no consequence would somehow compel them to do something few of us are inclined to do: slow down.
Scott Kelley’s idea, being so counter-intuitive, proved a marketing nightmare. No town official anywhere in America was placing orders. So he did the only thing he could - he sent out free samples for testing. One ended up in his own neighbourhood.
‘I remember driving by it,’ he said. ‘And I slowed down. I knew there was no camera in it taking my picture. Yet I slowed down. I just went, “Wow! This really does work!”’
In test after test the results came back the same. People did slow down - by an average of 14 per cent. And they stayed slowed down for miles down the road.
‘So why do they work?’ I asked Scott.
