Social media helped Donald Trump take the White House. Silicon Valley won’t let it happen again...
In the 1962 Japanese sci-fi classic King Kong vs Godzilla, the two giant monsters fight to a stalemate atop Mount Fuji. I have been wondering for some time when the two giants of American social media would square up for what promises to be a comparably brutal battle. Finally, it began last month - and where else but on Twitter?
‘Facebook was always anti-Trump,’ tweeted the President of the United States on 27 September.
Mark Zuckerberg shot back hours later (on Facebook, of course): ‘Trump says Facebook is against him. Liberals say we helped Trump. Both sides are upset about ideas and content they don’t like. That’s what running a platform for all ideas looks like.’
A platform for all ideas? Well, maybe. Others see Facebook differently. As Zuckerberg’s response to Trump acknowledged, the President is not alone in criticising him. The various inquiries into Russian meddling in the 2016 presidential election are turning up much that is awkward, notably that Russia bought around 3,000 Facebook ads designed to spread politically divisive posts to Americans before and after the election, as well as to promote inflammatory political protests on issues such as Muslim immigration.
It may be too big a stretch to claim that Russian Facebook ads swung the election in Trump’s favour. But it seems plausible that his campaign’s use of social media, particularly Facebook, gave it a vital edge that compensated for its financial disadvantage relative to Hillary Clinton’s campaign. On that, if on nothing else, I suspect Steve Bannon and Clinton would agree. ‘Facebook is now the largest news platform in the world,’ Clinton writes in her election postmortem. ‘With that awesome power comes great responsibility.’
Awesome power, yes. At the end of June, the number of active Facebook users (people who visit the site at least once a month) passed the two billion mark. WhatsApp, Messenger and Instagram — all owned by Facebook — have three billion users altogether, though no doubt there is much overlap. Two- thirds of American adults are on Facebook and 45 per cent get their news from it. More than half the UK population access Facebook at least once a month. The average user is on the site for 1/16 of every day.
But great responsibility? In the wake of the Las Vegas massacre, Facebook briefly featured a bogus story that the shooter had ‘Trump-hating’ views. A fake page claimed responsibility for the attack on behalf of the far-left Antifa movement, saying the goal had been to kill ‘Trump-supporting fascist dogs’.
Last month, the non-profit investigative news site ProPublica revealed that Facebook’s online ad tools had helped advertisers to target self-described ‘Jew haters’ or people who had used phrases such as ‘how to burn Jews’. In the words of Facebook’s chief operating officer Sheryl Sandberg: ‘The fact that hateful terms were even offered as options was totally inappropriate and a fail on our part.’ Facebook ‘never intended or anticipated this functionality being used this way’.
What Facebook intended and how Facebook is used turn out to be very different. The company’s motto used to be: ‘Make the world more open and connected.’ It’s no longer quite so simple.
‘For most of the existence of the company, this idea of connecting the world has not been a controversial thing,’ Zuckerberg recently said. ‘Something changed.’
What has changed is that the world has belatedly woken up to realities about social networks that were already obvious to anyone familiar with history and network science.
For most of history, it is true, hierarchies have tended to dominate distributed networks. However, there are historical precedents for technological change leading to enhanced connectedness that empowers social networks and weakens hierarchies.
The first began exactly 500 years ago, when Martin Luther launched his campaign for reform of the Roman Catholic church. Had it not been for the printing press, Luther would have been just another obscure heretic and might well have ended his life in the flames of the stake. But Gutenberg’s innovation enabled Luther’s message to ‘go viral’ — as we would now say — and it spread with remarkable speed throughout Germany and then across north-western Europe.
Luther was as much of a utopian as the pioneers of Silicon Valley in our own time. In his mind, the Reformation would create a powerful new network of pious Christians, all enabled to read the Bible in the vernacular and to establish more direct relationships with God than the indirect ones mediated by a corrupt ecclesiastical hierarchy. The vision of St Peter of a ‘priesthood of all believers’ would finally be realised.
But the true upshot of the Reformation was not harmony but polarisation and conflict. Not everyone was inspired by Luther’s message. Some sought to go further than him. Others reacted violently against the proposed reforms. The Counter-Reformation adopted the Protestants’ novel techniques of propagation and deployed them against the heretics.
Yet it proved impossible to destroy Protestant networks, even with mass executions and hideously cruel torture. If anything, persecution promoted radicalisation. Meanwhile, the constantly growing network of printed words proved itself as ready to spread madness as holiness. The witch craze of the 17th century was a classic example of a monster meme, claiming innocent lives from Scotland to Salem, Massachusetts.
There are three big differences between now and then.
First, today’s social networks are vastly bigger, faster and more widespread than those of the early modern era.
Secondly, whereas the printing press was a truly decentralised technology — Johannes Gutenberg was no Bill Gates — the ownership of today’s IT infrastructure is concentrated in remarkably few hands.
Finally, our networked age began by disrupting markets and later politics; only one religion, Islam, has been significantly affected.
But the similarities are nevertheless striking. Now, as then, newly empowered networks have led to polarisation, not harmony. Now, as then, the networks have acted as a transmission mechanism for all kinds of manias and panics as well as truth and beauty. And now, as then, the networks have eroded territorial sovereignty, weakening the established structures of political authority.
The US government sought to harness the power of social networks when the National Security Agency co-opted the big technology companies into its PRISM programme of mass domestic and foreign surveillance. But the new networks did not easily integrate into old power structures. Globally disseminated leaks, courtesy of Edward Snowden and Julian Assange, exposed PRISM, while a new kind of populist politics flourished on social media.
A defining feature of social networks (as in the Reformation) is their tendency to divide rather than unite. Recent research on American blogs and Twitter reveals a similar pattern: the emergence of two self-segregated ideological communities, one liberal, the other conservative. Just as birds of a feather flock together (network geeks call it ‘homophily’) so Twitter users retweet within their political clusters. One study found that with tweets on hot-button political topics (such as gun control, same-sex marriage and climate change), the use of emotional words increases their diffusion by a factor of 20 per cent for each additional word. Ever wondered why tweets are full of expletives? Now you know.
The presidential election of 2016 was a tale of many networks. By going viral through a largely self-organised network, Trump beat Clinton’s old-school, hierarchically structured campaign, which poured money into antiquated channels like local television. Isis contributed to the febrile atmosphere with its worst attack in North America (in Orlando in June last year), prompting Trump’s populist (and popular) promise of a ‘Muslim ban’. But the Trump network had itself been penetrated by the Russian intelligence network. Trump’s campaign and, to a much smaller extent, the Russians both used Facebook and Twitter as tools to discredit his opponent and discourage potential Democratic voters.
Make no mistake: 2016 will never happen again. Silicon Valley hates Trump for too many reasons to count. The most important are his stance on immigration (on which the Valley depends for its supply of skilled software engineers) and Big Tech’s need to ‘virtue-signal’ to its most valued user demographic: the young and affluent. They lean left. So does the otherwise capitalist Valley.
The political consequences were not immediately obvious, unless you were paying close attention, but after the Charlottesville clashes between white supremacists, neo-Nazis and their various left-wing opponents, they were there for all to see. Matthew Prince, CEO of the internet service provider Cloudflare, described what happened: ‘Literally, I woke up in a bad mood and decided someone shouldn’t be allowed on the internet.’ On the basis that ‘the people behind the Daily Stormer are assholes’, he denied their fascistic website access to the worldwide web. As Prince himself rightly observed: ‘No one should have that power. We need to have a discussion around this with clear rules and clear frameworks. My whims and those of Jeff [Bezos] and Larry [Page] and Satya [Nadella] and Mark [Zuckerberg] shouldn’t be what determines what should be online.’ Yet that discussion has barely begun. And until it happens, it will indeed be they who decide who is allowed on the internet.
This goes to the heart of the matter. According to Zuckerberg, Facebook is ‘a tech company, not a media company… We build the tools; we do not produce any content’. Yet in practice, according to a recent Reuters investigation, ‘an elite group of at least five senior executives regularly directs content policy and makes editorial judgment calls.’ In the words of Espen Egil Hansen, the editor-in-chief of the Norwegian newspaper Aftenposten, Zuckerberg is now ‘the world’s most powerful editor’.
It is not only neo-Nazi sites that find themselves on the online equivalent of the newsroom spike. Twitter has recently rejected paid-for tweets from the Center for Immigration Studies (CIS) on the grounds of ‘Hate’. These tweets were hardly excerpts from Mein Kampf: for example: ‘The fiscal cost created by illegal immigrants of $746.3bn compares to a total cost of deportation of $124.1bn.’ In the words of CIS director Mark Krikorian, ‘The internet is now a utility more important than phones or cable TV. If people can be denied access to it based on the content of their ideas and speech (rather than specific illegal acts), why not make phone service contingent on your political views? Or mail delivery?’
Google recently revealed that it is using machine learning to document ‘hate crimes and events’ in America. Among their partners in this effort is the notorious Southern Poverty Law Center (SPLC), which maintains a list of ‘anti-Muslim extremists’ — including my wife, Ayaan Hirsi Ali, and the British liberal Muslim Maajid Nawaz — but no list whatsoever of Muslim extremists.
‘YouTube doesn’t allow hate speech or content that promotes or incites violence,’ declared a recent message to YouTube content creators. But who decides what is ‘hate speech’? The phrase has become the 21st-century equivalent of ‘heresy’. It’s what you call something before you proscribe it.
Silicon Valley insists it is home to neutral network platforms. This is no longer credible. Facebook alone has, without quite meaning to, evolved into the most powerful publisher in the history of the world. Zuckerberg is William Randolph Hearst to the power of ten.
So what to do? Left-leaning Democrats have an answer: revive the progressive interpretation of anti-trust policy and break up the internet monopolies. Superficially, they have a case. Amazon controls 65 per cent of all online new book sales. Google’s market share of online search is 87 per cent in the US. In mobile social networking, Facebook and its subsidiaries control 75 per cent of the American market.
Yet who seriously cares what the hipster anti-trust types say? Silicon Valley is a huge donor to the Democrats. Why would they make life difficult for Big Tech when it so openly leans left? The real question is when Republicans (and not just the President) are going to wake up to the threat they now face.
Two big battles are looming: one on the question of net neutrality (the principle that all bits of data should be treated alike, regardless of their content or value), the other on the 1996 Communications and Decency Act, which allows tech firms exemption from liability for content that appears on their platforms. A group of senators led by Rob Portman has started the ball rolling by seeking to impose liability on companies that knowingly facilitate sex trafficking on their platforms. The initial response of the Internet Association, a trade group that is essentially a mouthpiece for the Valley, was revealing: ‘The entire internet industry wants to end human trafficking,’ it said. ‘But there are ways to do this without amending a law foundational to legitimate internet services.’ Last month, however, the IA conceded the need for ‘targeted amendments’.
This battle is only just beginning, but its outcome could be decisive in both the 2018 midterm elections to Congress and the 2020 presidential race. The regulatory status quo is not only highly favourable to Silicon Valley. It could also prove highly unfavourable to Republican candidates — though (so far as I could tell on a recent visit to Washington) the penny has not yet dropped with lawmakers who are accustomed to talking only about deregulation, not regulation.
In many ways, what we are about to witness will be a classic struggle between new networks and established hierarchies. Like King Kong’s epic slugfest with Godzilla, however, it’s far from easy to predict which side will prevail - or how much collateral damage they will both inflict on American democracy. In the old Godzilla movies, after all, the one predictable thing is that Tokyo always gets destroyed.