In addition to washing your hands while singing the first two verses of The Internationale, it might be a good time also to clean out your Twitter feed. According to a recent report of a research study by Carnegie Mellon University’s Center for Computational Analysis of Social and Organizational Systems, about 45% of the false narratives about Covid-19 on Twitter are sent by bots.
The study examined more than 100 false Covid narratives (including the 5G conspiracy theories) pushed in over 200m tweets since January. If you’re a reader of this newspaper, the likelihood is that you never saw any of these. But that’s because you are – like me – cheerfully encased in your own filter bubble. I write with feeling on this matter, because on the morning after the Brexit referendum I went through the list of about 800 people whom I follow on Twitter, and I could not locate a single one who seemed to have been in favour of Brexit in the run-up to the vote. The shock felt by them after the vote was palpable. But it was also a salutary reminder that anyone who uses social media lives in a digital echo chamber.
To those who study computational propaganda, the findings will come as no surprise. Philip Howard’s group at the Oxford Internet Institute has been doing great work in this field for several years. It would be nice to think that the mainstream media might have learned a thing or two from this work. If it had, it might have learned that just as you cannot believe anything you read in a British tabloid newspaper, you should never regard what’s “trending” on Twitter as anything other than an artefact of robotic amplification.
But bots are not the only problem facing Twitter. An equally difficult challenge is that presented by Donald J Trump, currently the president of the US. Twitter was a medium made for Trump, and he has exploited it masterfully ever since he contemplated running for public office. Although his campaign adroitly exploited targeted advertising on Facebook, Twitter was always how Trump directly communicated with his base. He currently has 80.3 million followers and has tweeted over 52,000 times. And while nobody was surprised that he campaigned using Twitter, nobody in the liberal (or even the conservative) establishment dreamed that, once elected, he would govern by tweet. Which is why the ingenious @RealPressSecBot bot that transforms Trump tweets into the proper presidential White House statement format is both such a good joke, and so depressing.
The fact that Twitter has allowed Trump to abuse the service by spreading lies, libels, conspiracy theories, nonsense and racist abuse continuously since he was elected has infuriated many people. Anyone else doing that kind of stuff would long ago have been banned from the platform. But to date, the company’s executives have shrunk from doing anything about the tweeter-in-chief. They have two reasons for their forbearance, one good and one bad. The good reason is fear for their lives. There are an estimated 392m unregistered firearms in the US, and my hunch is that banning Trump would eventually result in a Charlie Hebdo-type attack on the company by alt-right fanatics. The bad reason is that Trump is good for “user engagement”, which is the holy grail of all social media platforms.
This week, though, they took a baby step towards trying to rein him in. After he tweeted: “There is NO WAY (ZERO!) that Mail-In Ballots will be anything less than substantially fraudulent. Mail boxes will be robbed, ballots will be forged & even illegally printed out & fraudulently signed. The Governor of California is sending Ballots to millions of people, anyone.....” a label in blue lettering appeared underneath it urging people to “get the facts” about voting by mail. The links led to a CNN story that said Trump’s claims were unsubstantiated and to a list of bullet points that Twitter had compiled rebutting the inaccuracies.
In any normally run polity, this would have been an unremarkable step for a platform to take. In the crazily polarised society that the US has become, though, it will be seen by Trump and his fans as an affront to his Supreme Leadership. Not quite on the Charlie Hebdo scale of publishing cartoons of the Prophet, perhaps, but an affront nonetheless. At best, legions of bots will now be enlisted to pump abuse at the Twitter bosses and pump more misinformation about postal voting to ordinary users of the platform. More dangerous – not only to Twitter but to the business models of all social-media companies – would be a move by a vengeful Trump to use the nuclear option’ of removing the shelter provided by Section 230 of the 1996 Telecommunications Act which frees online platforms from legal liability for what users post on their platforms. Ironically, though, if Trump did deploy that weapon, then the rest of us would be free to sue Twitter for what the president tweets!
What I’ve been reading
Bringing it all back home
“How to fix globalisation – for Detroit, not Davos” is a really fascinating interview on the American Interest site with Lawrence Summers, a former US treasury secretary.
A time of loss
There’s an unexpectedly moving post, called “What We Leave Behind”, about the Covid-19 fallout by Scott Galloway on his blog.
No more robot overlords
In a New York Times article, Ben Schneiderman lays out his view that we need machines to augment human capabilities rather than replace us.