Mr. Hank Green, a prominent American Tiktoker, posted a viral YouTube video in the second week of October 2020 urging his vast number of fans and followers to watch a town hall event featuring Joe Biden and aired on ABC News, on multiple devices like their televisions, laptops and mobile phones. This ABC-News event was one of two parallel events being held on the same day and at the same time, the other being that of President Donald Trump on the NBC network, just ahead of the presidential election. Following his diktat, hundreds of thousands of young Tiktokers, already angry with the Trump administration for banning any transactions on the Tiktok app, watched the Biden event on multiple devices simultaneously to push the Democratic presidential candidate’s town hall ratings higher than those of Trump by almost a million. Per se, these TV ratings on their own did not mean anything in terms of the actual voting indication of Americans, but they did have an indirect impact of much greater proportions. Following the two events, various newspapers came out with screaming headlines declaring the Biden event a grand success and that of Trump an utter failure based on the ratings gap between them. Many op-ed writers in the vast jungle of US print media then wrote reams of newspaper columns on Trump’s ascendancy to the throne of the most-hated president in US history, citing the astronomical differences in the TV ratings of the parallel events of both candidates as the latest example of his increasing unpopularity. This is an example of how virtual communities can impact real-life events.
For many months prior to these TV events, hundreds of national opinion polls conducted by dozens of pollsters projected the sitting president to be trailing his Democratic counterpart by increasing gaps with every passing week. By October, some of the prominent national polls conducted by big media houses like New York Times, Washington Post, ABC News, NBC News and The Wall Street Journal had begun to show a blatant bias in their polling numbers by projecting a lead of over 12% to Biden, despite knowing that such leads were data-metrically improbable in the current high partisan scenario – the last time a US presidential election saw a double-digit lead was back in 1984 when American society was not yet divided so deeply along ideological/party lines. Other, highly rated independent polls and pollsters like Emerson, IBD/TIPP and Zogby associates, along with the more Republican-leaning Trafalgar (which was the most accurate pollster of 2016) and Rasmussen, who predicted a much tighter race were unanimously blocked out by the narrative nodes. Facebook and Twitter curation algorithms ensured that only the polls by big media houses that predicted mammoth leads for the Democratic candidate were widely circulated, whereas other polls were effectively lost in the rubble of cyberspace.
During the third week of the same month, on 15th October, the New York Post, the third-largest selling newspaper in the US published the now-famous “Hunter-Biden emails” purporting his alleged corrupt activities. Twitter and Facebook immediately banned the act of posting any of those newspaper articles on its platforms. Search-engine giant Google too was not far behind as most search queries related to Hunter-Biden were effectively stifled by down-ranking the New York Post article and instead ensuring other counterfactual news content to dominate the front pages. The algorithms deployed by big-tech companies had moved beyond from “effective curation” of content to “actual blocking” of information.
Something even more sinister with a far-reaching impact transpired in the technological universe at the fag-end of the US election season. On 29th October 2020, some 4 days before the election, the “go out to vote” reminder message disappeared from Google’s homepage across the US. In the preceding week, the American Institute of Behavioural Research and Technology had “warned” that Google’s homepage message was targeted only towards liberal voters while completely ignoring the conservative demographic. Essentially, Google had used the trillions of data points available to it on American voters to classify them into “liberal” and “conservative” segments. Then, as the Behavioural Research paper on these segment of voters showed, Google only sent out reminder messages to targeted “liberal” voters when they opened their homepages on laptops, computers and mobile devices, whereas voters of “conservative” persuasion were effectively blocked out from receiving this reminder message on their respective homepages. By a quirk of “fate” or astronomical coincidence, on 29th October, the very day that the study by the American Institute of Behavioural Research and Technology was to be published, Google conveniently “switched off” its public-service message!
Dr Robert Epstein, a renowned mass behavioural scientist who headed this particular study has concluded that Google’s search engine algorithms alone could have potentially shifted anywhere between 3 to 15 million American voters by what is known as “suggestive persuasion” methods. In fact, Dr Epstein’s long-running research on “Ephemeral Experiences,” which form the basis on which Google alters consumer preferences by altering search engine results, is an important clue on how big tech can subtly shift mass behaviour in small percentages over time. Dr Epstein describes Ephemeral Experiences as “those experiences we have online that involve very fleeting content that impacts us, disappears, and is gone forever and can’t be analysed because it disappears.”
In the end, when the election results did come, they conclusively proved the utter falsehood of the narrative of an impending “blue wave” (referred to as the Democratic party landslide). In fact, Donald Trump lost the election by a few thousand “disputed” mail-in ballots in four key battleground states of Arizona, Georgia, Pennsylvania and Wisconsin (as of writing this piece, Trump is yet to concede defeat and his campaign team has filed dozens of lawsuits against hundreds of thousands of disputed mail-in ballots). The 2020 US presidential election will go down in history as that inflection point when big tech had amassed enough power to tilt the scales of global democracies.
In India however, there has hardly been a ripple, in terms of governance or policy changes or even the discourse surrounding the increasing power of social media and big tech corporations. Many politicians and political commentators make valuable but specious arguments that India is too vast a country with too many digitally illiterate people for internet-based technologies to have any impact on their voting decisions. Indeed, very few people can comprehend the staggeringly deep impact of technology on our societies. The reality though is quite stark. We are closer to a big-tech takeover of the global order than we have been ever before; it is now a matter of years and not decades. If all of this sounds too alarmist to you, then let us consider some recent data from the algorithmic world of social media to understand these subtle shifts and the resultant sea changes in mass behavioural patterns.
A promising young Bollywood actor, one of the few outsiders to succeed in an industry filled with nepotistic successes, committed suicide in June and gave rise to a universe of social media algorithms that thrive on conspiracy theories of all kinds and sizes. The entire audience pool of the late Sushant Singh Rajput’s films put together could be deduced to be not more than six crore (60 million) Indians – it is a not so complex an exercise to create a regression algorithm that combines the box-office footfalls, television impressions and downloads of all his films along with accounting for redundancies among different pools of audience subsets. Now compare this with the number of people who have consumed at least one or other form of conspiracy theory that surrounds his death and you will see the staggering difference. Conservative estimates show that not less than 30 crore (300 million) Indians have consumed at least one conspiracy theory or the other surrounding Mr. Rajput’s death, about five times more than the entire audience for all his movies put together.
Welcome to the world of screen looping, where information travels and gets amplified from one screen to the other seamlessly. There are, for instance, less than a million active Twitter users in India – a microscopically small number within a veritable population of nearly 1.4 billion. Yet, the influence that Twitter has on narrative nodes is many, many magnitudes higher. This small pool of a million or so active Twitter users act as the inner circle of a large ripple of information cascade. News, opinions and information that originate in real-time on Twitter get transmitted to other screens within one or two short hops and then reaches the wider masses either through other more massy social-media platforms like Facebook/WhatsApp or through vernacular news and media outlets. This is how the Sushant Singh Rajput suicide case (known as the SSR case in the online parlance) acquired a life of its own.
Majority of the consumers of the news and opinions surrounding the SSR case belonged to what is euphemistically known as the “youth” segment. It is this segment of voters who have been historically lauded as being the harbingers of anti-establishment sentiments in any geography. For instance, in the mid-70’s, it was the youth of Gujarat and Bihar who first turned against Indira Gandhi and became the trailblazers of the JP revolution across India. Similarly, the anti-corruption movement pioneered by Baba Ramdev, Sri Sri Ravishankar and Anna Hazare during the UPA-2 regime that ultimately led to the virtual decimation of Congress was mostly a youth-powered crusade against the excesses of the Sonia-Manmohan regime. Since the advent of Narendra Modi on the national scene in 2014, many liberal and left entities in India have tried repeatedly to create unrest among the youth of India and been disappointed time and again. Whether it was the Rohit Vemula suicide case or the various protests in university campuses (mostly flagged from JNU) or even the anti-CAA violence, every attempt to create disenchantment among the young and restless in India has met with ultimate failure.
It was in this context that the Bihar election, held in November, proved to be a strong contraindication for the first time since 2014. As per our analysis of various focus groups in Bihar and the third-party polling data from the state, we conclude with utmost certainty that 70% of the Bihari youth – nearly 3 out of 4 voters belonging to the 18 to 30 age group – voted against the NDA in the assembly election, cutting across caste lines. This is almost a diametric turnaround when compared to the 2019 national elections when nearly 2 out of every 3 “young” voters (over 63%) had demonstrably voted for the Modi-led alliance in this same state (this author and his teams had executed deep-down surveys across India, including in Bihar, during the 2019 elections and had projected the elections with an accuracy level of 99.6%, weeks before the official results were announced).
Obviously, there were many factors at play here – the global pandemic-induced economic collapse and the relatively uncertain state of employment being one of the primary reasons – but the algorithmic impact on catalysing the anger among the young is the single most important factor that has been missed completely. Having had an ‘Ephemeral Experience’ of fleetingly consuming news and opinions on the SSR case, a vast majority of Bihari youth had strongly reinforced their feeling of an “unjust system” that cannot give them justice of any meaningful kind. It was such an anger against the perceived injustices that was fanned during the Bihar elections to create an overnight anti-establishment platform for RJD and Tejashwi Yadav, who despite being completely outnumbered and out-resourced on the ground put up a strong fight and came very close to winning this election.
These youth-vote shifts are important trend reversals that the BJP-Sangh leadership cannot brush under the carpet, for they portend the setting in of anti-incumbency which generally begins to manifest at the start of the second term of any government. For instance, even Indira Gandhi who had won a massive mandate in 1971 and had then won a war against Pakistan to liberate Bangladesh went on to face mass agitations within 2 years of governance and never recovered after that. Similarly, UPA 2 lost its way immediately after winning an enhanced mandate in 2009. In fact, no national government has won a third successive term since the one headed by Jawaharlal Nehru in 1962. Furthermore, a global pandemic of unprecedented proportions has created a whole new dynamic in the political economies of the world today – a dynamic that can last for many years and give rise to hitherto unforeseen problems for Modi and the BJP.
In this shifting historic background, big tech corporations are operating toward their own set of goals and objectives. Mark Zuckerberg had famously offered a philosophy of history in his 2017 blog (as a reaction to Trump’s victory in the 2016 US elections), a history that according to him has moved in an arc from tribes to villages to cities to nation states – and finally to something beyond. “Today we are close to taking our next step,” he had written about humanity as a global community built on the “social infrastructure” provided by his Facebook-enabled corporations. His argument was ultimately developed on a rather farfetched ambition of big tech replacing the state as a primary builder, organizer and operator of human communities.
Today, the notion does not seem so farfetched after all.
Already, most of us humans are nothing but streams of data that are being constantly mined and analysed by big tech for the best possible predictive deployment – today, a Google, for example, knows not only far more about you and me than what any state or government does, but in many cases it knows us more intimately than do our friends or closest relatives.
It is this new world order that big tech is spawning, which is at the heart of the most important battle of our times. To be sure, over the previous decade, when big tech (especially, social media platforms) began their first cycle of expansion, they had managed to remain fairly neutral in their approach. As a result, they became enablers of hitherto lost voices of conservative nationalism all over the world, and consequently political systems changed drastically from an overcrowded left-liberal space to a more nationalist trend. In India, before the advent of Modi in 2014, the entire political spectrum was filled by parties that belonged to the left-of-centre space, often even prefixing the terms “secularism”, “socialism” or “samajwad” to their anachronistic existence. Even the two terms of Vajpayee during the late-90’s were enabled by various smaller NDA parties that acted as secular-socialist enforcers of the coalition government. Once these social media and big tech platforms achieved scale by attracting billions of users, they began to shift from their neutral approach and instead ascended to the role of judge and arbiter for all information by imposing ideological preconditions.
Throughout their initial years, Twitter’s policy was built on the philosophy of its co-founder, Biz Stone, who had summed it up in the famous 2011 blog post as, “The tweets must flow.” Unless the tweets were explicitly spam or broke specific laws, the idea was to let the good tweets be the antidote to bad tweets and eventually balance overall information flow towards positivity under its own steam. Sometime in October 2015 things began to change radically, coinciding with the return, after a brief hiatus, of Jack Dorsey as the head of Twitter. It was around this time that there was a global backlash by the liberal elite who could not comprehend the fact that they no longer controlled the narrative nodes of global politics. Whether it was Brexit in Europe or the rise of Donald Trump in the US or the advent of Modi in India, the left-liberal-dominated traditional media-intellectual ecosystem found itself overwhelmingly outnumbered and outsmarted by ordinary people who refused to accept the wisdom handed down by these traditional “elite” talking heads. Nonplussed by the events unfolding in front of them, the liberal elite began to denigrate these new political realities as “the post-truth world” cultivated by “WhatsApp universities” and other such disparaging terms and began to harangue the big tech corporations to change the direction of the discourse.
Silicon Valley, already home to some of the most left-liberal voices in the world, reacted swiftly and with alacrity. For instance, after Jack Dorsey’s return to Twitter, Vijaya Gadde, the legal head of the company, was given a complete free hand to alter the policy from a free-for-all platform to a content-moderated platform. The shift seemed subtle at the outset, but had monumental impact on the social media universe. Ms. Gadde, an Indian-origin expat who moved to Texas at the ripe young age of two, had some life-altering experiences during childhood when her chemical engineer father lost his job and had to get permission from a local Ku Klux Klan leader for collecting door-to-door insurance premiums. These experiences shaped her outlook in life and she has now become somewhat of a liberal fundamentalist who is antagonistic of the supposedly “majority” communities anywhere and sees herself as the voice of the “marginalised minorities” of the world. When Vijaya Gadde and Jack Dorsey visited India in late 2018, they had an open discussion with various “activists” and what has been loosely termed as the “urban naxal” gang. Since that visit, Twitter’s policy in India has undergone a dramatic transformation.
Such changes have not been limited to Twitter per se. Other big tech corporations too had been working towards affecting similar makeovers. Facebook began its massive content-moderation drive around the same time, which accelerated due to the liberal anger after Hillary Clinton’s ‘shock’ defeat at the hands of Donald Trump. Today Facebook reportedly has 15000 “content contractors” across the globe who are tasked with managing content that is posted on their platform by all the users. Google went on a more technological path to curb the narrative changes owing to its gigantic data repositories. Down-ranking algorithms for selective search results to alter narratives became the vehicle of choice for Silicon Valley.
As data scientist Cathy O’Neil explains succinctly in her book Weapons of Math Destruction, “Algorithms are opinions embedded in code… algorithms are not objective at all.” Algorithms and AI deployed by big tech are not neutral, they are shaped by ultra-left-liberal entities like Vijaya Gadde occupying leadership positions and deploying a combination of tools at their disposal like banning users, curating content and down-ranking information nodes, all targeted against Right Wing Nationalist voices to upend global democracies. In India, many nationalist voices have been banned outright by both Facebook and Twitter, which is an easier bias to discern. What is more subtle is the way curation algorithms work by limiting the reach of Tweets and Facebook posts. For instance, many hundreds of pro-Hindu, Nationalist Twitter handles (including this author’s) have been “algorithmically frozen” with no new followers getting added for over a year and limiting their retweets arbitrarily. As a result, a vast majority of “low information” ordinary users only get one-sided information overdrive despite perceiving these tech platforms as being the neutral voices of the common man.
This is a far bigger problem than what many people can comprehend. Most ordinary Indians have little trust in the traditional mainstream media, but perceive big tech and social media as decentralised democratic platforms that reflect the voices of ordinary people. Since these technology companies are no longer neutral in their approach, they are ensuring that only ideologically left-liberal aligned voices get magnified while those opposed to that view are marginalised to their own echo chambers. Over a period of time, most Indians, especially the young, will be systematically exposed to ‘Ephemeral Experiences’ that aim at reinforcing a particular ideology. It is a world of “vastly expanded private mechanisms of social control, based on rendering human behaviour fully predictable,” in the words of Shoshana Zuboff, the author of the much-acclaimed work Surveillance Capitalism.
As media theorist Steven Johnson hypothesised, “although internet was built on open protocols that created such ground-breaking tools as the email, search, and browsing… what the inventors of these protocols (initially) failed to understand was that they were building a community and not just a machine.” It is this dichotomy of a machine and the community that has enabled the centralization of internet into the domain of a handful of technology companies. Just like a simple calculator could acquire virtually limitless capabilities through an internet connection by transforming itself into a sophisticated computer, these community-machines began to connect to outside functions in the real world within a few years. The question then is, what is Democratic Singularity in this scenario? One begins to approach the state of Democratic Singularity as this community-machine begins to acquire increasing number of functions of electoral systems – it is at this moment when these community-machines are no longer virtual systems distinct from reality but instead become the controlling protocols for governing democratic exercises in the real world that Democratic Singularity would have commenced.
The big tech community-machines have already tasted blood in the 2020 presidential elections. A global pandemic that has enforced severe limitations on physical contact within human communities has meant that acceleration of tech-powered interaction services has increased by many magnitudes over the last 1 year. We are now approaching Democratic Singularity at a much faster pace than just a year ago. Political parties, governments and, ultimately, the state itself is severely limited in its capacity to either understand or harness the sheer volumes of data that are being produced with each passing minute. It was never a question of “if” and always a question of “when,” now that “when” is clearly visible on the horizon as the world’s richest democracy grapples with an election that was not only guided and manipulated by big tech, but also by targeted avoidance of in-person voting exercises which will now become the harbinger of the future of “remote voting technologies” wherein big tech can further enhance its role as the judge and the arbiter of individual choice.
In India, after many centuries of struggle, the Hindu political renaissance is finally being achieved just in the last few years. One can imagine the scale of subjugation of Hindus that it took us nothing less than 500 years – yes, half a millennium! – to acquire a small piece of land to rebuild a temple for lord Rama at his birthplace in Ayodhya. The fate of political Hindutva could soon become one of the earliest casualties as we approach Democratic Singularity. Modi, BJP and the Sangh do not have any political opposition today in India. Opposition is irrelevant. As it turns out, opposition is not needed anymore. Take the case of the Bihar elections, for instance. There was no opposition whatsoever; if anything, Tejashwi Yadav was the dream opposition that one could only hope for, he is a tainted dynast, has little political infrastructure and had remained absent from the state’s political firmament until a few weeks before the election. And yet, the junior Yadav lost the chance of becoming the chief minister by a few thousand votes. Surprisingly, the Joe Biden case is quite similar. Here was a presidential candidate who never campaigned, almost never set foot outside of his basement and evoked very little enthusiasm among the voters, and yet somehow managed to obtain the highest number of votes in US election history and became president by getting a few thousand additional mail-in votes in half-a-dozen battleground counties.
The algorithms and the AI deployed by big tech will keep getting stronger and more precise in time as more data gets fed into the system and soon the spectre of Democratic Singularity will get to become an irreversible process. How India reacts to this dynamic is the primary question of this decade, which will reflect on our political economy and ultimately decide the fate of our democracy. Although, a quick-fix solution of banning many of these platforms may sound good in theory, it will be highly counterproductive in the long run as technologies will metamorphose and keep entering through backdoor channels. India’s reaction will have to therefore depend on a multi-pronged strategy of policy intervention, coercion, data-tech adoption and course correction. As always, this reaction of India will be shaped by our government and our political leadership.
The question then arises whether the Sangh and BJP leadership have the time, the willingness and the wherewithal to emerge victorious in the smartest battle they have fought to date?
[This article was first published in the Organizer, you can read it here ]
Hi