Oct 17, 2019

Social Media: Ground Zero for the War on Ideas

Julian Glassford

This time last year I cautioned, in a commentary for the libertarian e-zine Spiked, that Twitter’s leadership had finally sold-out on free speech. Much to the consternation of right-minded users of the social network, Jack Dorsey & co’s outfit had begun to surreptitiously stem the free flow of dialogue, ideas, and information via a wave of so-called ‘shadow bans’.

Undeterred by such criticism, sadly the platform has doubled down on creeping censorship and mass deplatforming in the time since. I, myself, recently discovered that many of my tweets were subject to (inexplicable) ‘deboosting’ i.e. the process by which social media comments are deliberately hidden or otherwise obscured by online platforms.

On Twitter the process entails replies to tweets being hidden behind a ‘show more replies’ link buried at the bottom of the list of unadulterated replies below any given post. Consistent with reported shadow bans, the network is doing this without warning, explanation, or effective means of appeal. Affected users are simply left, to languish, in the dark.

In my case, the platform’s support service did not bother to respond or take any apparent action in relation to contact queries on this matter, and no ‘deboost’ information popped up when I searched their help portal. And so it seemed my further contributions were destined to be auto-filtered out of public consciousness, on the sly.

Naturally, this had the not unpredictable effect of disincentivising and generally disenabling my continued participation in broad-based political discourse. Reflecting on this, and seeing no reason to maintain a ghost of a Twitter presence, I have for the time being deactivated my personal account. Mission accomplished, Jack?

In acting in this way Twitter is, in effect, silently going about working to help institute a two-tier system of online citizenry, i.e. ‘digital apartheid’, apparently in an attempt to insulate users from too much information (of the wrong type); and, by extension, perhaps also to minimise moderation costs and complexity. So much for the firm’s guiding values, purportedly centred on belief in: “free expression” and the power of “every voice … to impact the world”.

In an earlier analysis written for the progressive news site LeftFootForward, I explored some of the dangers posed by the run-away train that is (neoliberal) digital transformation.

In it, I observed that “certain major search engines and social media platforms have transformed into data swiping ‘attention merchants’, relentless marketeers, social engineers, and unelected arbiters of truth”. Per the above, this has now been further borne out by my own treatment.

In the absence of due clarification on the part of those responsible, I cannot discount the possibility that I may have been targeted precisely because I dared to undermine the Big Tech Gods in drawing attention to the unethical/unsustainable nature of various aspects of their newfound dominion. However, having investigated, it seems likely that users are being deboosted when they use particular (patterns of) taboo words or links. Users who do not follow all that many other accounts, or attempt to post replies to more than a small number of tweets from accounts with large numbers of followers, may be particularly vulnerable e.g. possibly (mistakenly) targeted by the algorithm as potential bots or ‘bad actors’.

So what? You might say. Really worth kicking up a fuss? You might add. Perfectly understandable responses, but such experiences form part of a wider trend that is in fact rather serious indeed. On a scale of one to proroguing Parliament for an extra week it is … well, let’s just say it’s off the scale. Why? Because, as I observed in writing recently to Jack & pals: “Twitter enjoys an ostensive monopoly of informational & ideas exchange, and in the political domain in particular”.

The comparison to shutting down Parliament may seem a little farfetched; but the stealthy and unjustified smothering of particular voices, in what has swiftly become arguably the most directly and immediately significant pool of public (political) discourse, represents an insidious threat to truly vibrant, healthy, and pluralistic tech-connected (political) systems.

Twitter is now the go-to common resource for most journalists, commentators, and reporting organisations (of virtually all kinds), as well as a large, vocal contingent of their audiences, critics, and subjects alike – including almost all politicians and a great many other public figures. Beyond this, it has become a powerful social barometer and unique tool for not just social networking and commercial marketing, but also non-commercial educating and influencing – across virtually every domain under the sun: from Accounting to Zoology.

Jack’s platform has – somewhat inadvertently – helped to deliver unprecedented improvements in the capacity of the average Joe to scrutinise, interrogate, debate, and remonstrate. For these transformative, emancipatory new powers to be selectively impaired, apparently without good cause, is not just unacceptable but also dangerous. As I have previously warned, it is at once inherently discriminatory, intolerant, repressive, and (hence), by extension, alienating, radicalising, and potentially destabilising.

History records that arbitrarily disempowering entire cohorts pushes them towards the fringes of (digital) society – enhancing the ‘echo chamber’ effect that we are told should concern us all – and hence promotes the (re-)establishment of relatively radical factions and forces. This may suit the narrow purposes of certain interests, for a while at least, but rarely works out well for anyone in the long run.

If the internet is a ‘global commons’, and it most certainly is, then conversational and ideational online ‘markets’, and the (natural) monopolies that have come to corner them, must by now surely be considered integral to this precious, rapidly expanding and evolving digital ecosystem; democratised digital assets in their own right. That being so, the next question has to be: are such increasingly pivotal e-tools and resources not worth protecting from in-house (as well as external) manipulation and abuse? What is good for the goose is good for the gander, after all.

It’s not just Twitter, either. Whistleblowers from Google and Facebook have also reached out to Project Veritas in recent months: exposing targeted human and algorithmic blacklisting and deboosting (mal)practice apparently mostly aimed at interlocutors assumed to be right-wing e.g. on the basis of (automated) syntax analysis. If the reports we are getting are to be believed then such corporations have evidently strayed far from their purportedly humble, diverse, and progressive origins, and are fast becoming a prejudicial and oppressive law unto themselves.

An unknown number of people and organisations are being ever so quietly and gently excluded from the (digital) Speakers Corner and town halls of the 21st century – apparently often on the basis of little more than rough indicators pertaining to politics and linguistics; both a strong function of one’s personal background, education, and lived experience. Our only apparent crime: believing we could be part of “what’s happening” (Twitter’s slogan) and also be ourselves. Is this not the very definition of bigotry? Would it not have Orwell turning in his grave?

In the continued absence of suitable checks and balances Big Tech is now injudiciously constraining – if not (yet) completely purging – public opinion, where and when it suits. Naturally, this runs contrary to modern, inclusive Western values of the post-enlightenment, and is arguably at odds with universal human rights including liberty, equality, dignity, and freedom of thought, speech, and expression.

What we are witnessing, however subtle and underreported, is nothing short of calculated social engineering and political interference on an undisclosed but conceivably industrial scale. Forget largely contrived #RussiaGate narratives, and the like, #MissingVoices is the real deal and it’s happening right now, and right under our noses.

To take my case as an example: I had only a small number of ‘followers’ but my replies – mostly to posts from popular Twitter accounts – had registered hundreds of thousands of impressions prior to the imposition of this curious deboosting impediment. Extrapolate those figures across just a hundred, or a thousand, and never mind a million critical voices, like mine, and the potential systemic impact is huge. And that, it would seem, is the point.

A brief sampling exercise reveals that up to one in three replies to randomly selected BBC news report tweets about all too common media talking points – like Brexit, Trump, and climate change – are being deboosted. Many of the affected accounts/posts appear contrarian/sceptical, e.g. of liberal mainstream narratives, if not consistently left or right-wing, or generally all that radical, nor clearly in breach of the platform’s rules. Indeed, it seems that even some of the UN’s replies to their own tweets are being hidden.

Hard to say quite what the agenda is here, or quite what has gone wrong at Twitter, in the absence of more comprehensive data collection and analysis – or Twitter shedding light on things themselves, but readers can draw their own conclusions. What we do know is that a large tranche of public opinion is being hidden away and, judging by noticeably reduced (political) network activity in recent weeks and months, it seems many have cottoned on and decided, like myself, not to bother trying to be part of the conversation.

As outlined in my earlier commentary on shadow bans, the timing of these acts of (mass) manipulation could scarcely be more conspicuous. It can hardly be a coincidence that these moves come “just as contemporary technological trends have moved the average consumer of news away from traditional, reliably on-message mainstream media outfits” and towards dynamic grass-roots, community based information sharing. With major national elections on the horizon both sides of the pond, this does not look at all innocent and nor does it bode well.

The truth is, the powers that be are losing the ‘information war’ on multiple fronts. From failed attempts to brush off health, safety, security, and energy consumption concerns over 5G, ‘smart’ IoT, and AI technologies, to the decreasing resonance of FUD-based (technocratic) propaganda, the ‘little people’ are waking up and smelling the “covfefe”; and we can’t have that. Shut it down!

Prominent politicians stateside have been rather slow to properly recognise Silicone Valley’s attempts to neuter the net, and have yet to do anything serious about it. That is, other than invite tech execs to appear before Congress only to essentially perjure themselves in having the audacity to claim, on the record, that their outfits are free of bias and do not seek to skew public perception.

By contrast, policymakers this side of the Atlantic have treated us to an Online Harms White Paper that gives censorious corporates the green light to get creative in keeping the (increasingly disaffected) peasantry from encountering too much in the way of (awkward) ‘disinformation’. The agenda, if not the evidence, appears to reflect concern that continued exposure to unvetted data might promote (disruptive) ‘wrongthink’ vs. (stable) ‘groupthink’. Ominously, the consultation document suggests that companies should make disputed content less visible.

Whatever happens at legislative and regulatory levels we, each and every one of us, have an important public duty to ensure that the (digital) commons may be used and enjoyed freely, openly, and on an equal basis by all rule and law abiding (digital) citizens.

Whether the solution is a full-blown digital transformation commission, more narrowly targeted regulation, and/or (the simple threat of) anti-trust interventions, what has become clear is that we urgently need to see more (open source) transparency, legitimacy, and public accountability on the part of major internet monopolies – before Big Tech fully transitions to Big Brother 2.0.