“Social media companies are not the arbiters of truth.”
These are apparently the words of Courtney Parella, a member of the Tangerine Amin’s campaign team overnight after Facebook finally grew a pair (1) and removed a post claiming that children were “almost immune” to the coronavirus. Twitter has also blocked it, according to The Guardian overnight. The only real shock is that it’s taken this long for the social media platforms to properly respond to what, in this case especially, is pretty much criminal negligence and irresponsibility on Trump’s part. Again.
But let’s leave aside Trump’s Olympian levels of asininity aside for just a moment and concentrate on those words again:
Social media companies are not the arbiters of truth.
There is quite a lot to unpack in that one sentence, mostly because, in Internet terms, they actually are. And that’s part of the problem.
They are the final arbiters of what appears on their platforms.
But that’s where it starts to get a bit murkier, mostly becasue what these platforms are selling is us, to advertisers. It’s why they are so keen to offer us personalised experiences, because that way they can more efficiently target market segments for those willing to pay for them. And they can charge more for it, because there’s a higher likelihood that the targeted market segment will deliver increased value for that prospective advertiser.
So to appeal to advertisers, the platforms have to show not just that they have lots of users, who continue to log in, but that those users engage. That engagement can take many forms. Lots of people post (“original”) things, but most don’t. Most of us more usually either comment upon or share the content of others. Those comments, and their tone, give the social media companies, and their advertisers, an increasingly detailed picture of us, the more we use their platforms.
Worse yet, traditional journalism sources are under pressure from the online world; they have moved to engage with social media platforms. This is partly to try and remain relevant to a significant tranche of users who are moving away from traditional media sources, and partly about finding sources of revenue to replace dwindling print circulations.(2)
So where does truth fit in? Well, mostly it doesn’t in this model, at its simplest. It’s not hugely important whether something is accurate or not, as long as people engage with it. And that’s where the problems begin.
Donald Trump’s Twitter account has 84.7m followers. That’s a lot of engagement (3). It certainly puts him comfortably in the top 10 most followed people on Twitter (4). That level of engagement from such a high profile figure puts a great deal of pressure on Twitter, not to fact check what he says, but to ensure applying sanctions does not disrupt the company’s user community, and thus its share price.
“He told me there were truth tellers and they needed to know what the truth so as to tell it. And there were liars and they needed to know what the truth was so they could change it or avoid it. And there were bullshitters who didn’t care about the truth at all. They weren’t interested. What they spoke about wasn’t the truth and it wasn’t lies, it was bullshit. All they were interested in was their own performance.”
(from The Secret Commonwealth, by Philip Pullman)
The same goes for Facebook. In recent months, both platforms have added extra resources to allow a limited degree of fact checking, and “challenges” to misinformation. But this is proving somewhat contentious. The algotithms governing such things are part of the assets of the company, and not freely available for the world to see. We cannot tell what creiteria are being applied to what posts, or even when. We can’t even reliably tell where there is any human intervention in that process. So they are not only the arbiters of their content, but also of the decision making process used to determine its accuracy.
Quite apart from that, trying to tell whether what is written on each page is “true” or not is horrifcally difficult. What we’ve found out very quickly in the Information Age is that a lot of people really aren’t that good at it.(5) And even the best are caught napping sometimes, we’re none of us perfect. At first, with spoof news sites like The Onion, this blurring was (and in many cases still is) funny, because although the format was the same as “serious” news, the content was clearly not(6). However, as the move to online media grew apace, the techniques to hook people into less reputable and verifable sources took their insidious and creeping current shape. And it’s everywhere, from the spurious clickbait(7) headlines on some news sites and the inevitable “listicles” (yes, i’m looking at you, Buzzfeed), to the externally sourced content used by others. Determining trust chains, provenance and trusted fact checking sources (hello, snopes.com) is time-consuming, and sometimes difficult. And many people don’t want to do it, either because they can't (8), or won’t, for whatever reason. Even if we could apply a faultless machne learning algorithm to determine “fact” from “non-fact”, it’s quite possible that many would simply choose to ignore it because it does not correpsond to their own internal picture of the world. A web of trust and verifiability is lovely idea, but until people (and especially children) are grounded from a young age in the skills of bullshit filtering, we are fighting a losing battle to hold the torrent back, and not get submerged in it. It’s not like people aren’t writing about it (such as here, and here), but fewer poeple than ever seem to want to listen. At least for now.
Revealing the truth in the way I’ve described it would not work. There are too many habits, ways of thought, institutions, that are committed to the way things are and always have been. The truth will be swept away at once. Instead we should delicately and subtly undermine the idea that truth and facts are possible in the first place. Once the people have become doubtful about the truth of anything, all kinds of things will be open to us.
(from The Secret Commonwealth, by Philip Pullman) (9)
This is one of the great growing pains of the information age. Back in mid 15th Century Europe, when Gutenburg first unveiled his new printing methods, it’s easy to think (as the now often repeated history tells us) that there was an immediate outpouring of new knowledge and art. Indeed there was, but it wasn’t as unrestricted as we like to think. There were gatekeepers, those who controlled access to the new medium (as always happens), and those people controlled the flow of ideas into the world. Those people were the publishers — people who owned the presses and the means of distributing books — as scribes and clerics had been to handwritten books and scripts. They were open to exactly the same presuures from commercial, political and religious sources as today.
The growth of social media has happened even more rapidly than the adoption of Gutenburg’s prinitng. It took around 50 years for the number of books in Europe to go from a few thousand to around 9 million. Around this time (1557) the University of Cambridge’s Library had less than 200 items. In comparison, the number of pages on the public web has grown from zero (10) in 1990 to somewhere in the region of 60 BILLION pages today. And this doesn’t satisfactorily measure the amount of content on each “page”, which includes all of that social graph data that the likes of Facebook and Twitter hold in their enormous repositories..
It’s those large corporations (Google, Twitter, Facebook), whose business models are focused on telling advertisers what we want to be shown, appealing to biases we already have, and preferences we have already expressed,who truly are the arbiters of truth. We have let that happen. We wanted convenience; we wanted speed; we wanted personal service. Now, we just have to be grateful when one of them actually decides a man that many of us know is liar and a snake oil salesman, is lying, so they tell us so.
(1) A pair of what? The choice is yours. Bollocks? Ovaries? Take your pick: I’m not being gender specific here.
(2) That is going to go on for some time, as some sources, like The Guardian have eschewed the subscription model used by others, such as The Times. It’s difficult to tell at this stage whether either model is going to sustain old media companies into the medium to long term.
(3) At time of writing. It’s also unlcear how many of those are “real” followers, and not just bots. But the bare number is still useful.
(4) I bet number 1 just eats away at Donnie’s withered soul every day…
(5) We all know someone who keeps posting the worst possible stuff. If you don’t, it’s probably you.
(7) A phenomenon that now has its own word, and we almost all know what it looks like.
(8) As a result of not being taught, or having sufficiently developed the critical skills to do it. That’s a whole other debate to be had, about the value of media studies, and of criticality, in the emerging 21st Century world. Lives quite literally depend upon it.
(9) I’ve used those two Pullman quotes elsewhere, in a review I wrote of The Secret Commonwealth, but they’re just so prescient, and so accurate, I couldn’t resist using them again to illustrate.
(10) In 1990, the World Wide Web was still an internal only project at CERN. The first public webpages did not launch until 1991.