By 2016-17, Facebook had designed a business model that aimed at maximising the ‘user engagement’, that is how much time users spent on their platform, and started collecting data on what they liked and shared. The more time users spent the more data Facebook collected, so their algorithms could then monetise the information by targeting their ads to specific users. As users spent more time with Facebook, Mark Zuckerburg was getting richer. Commercial success of this business model prompted Facebook to give their algorithms a single, no-holds barred, overarching goal ~ to increase user engagement at all costs, to the exclusion of every other consideration.
By analysing data on millions of users, the Artificial intelligence (AI)-based self-learning algorithms discovered quickly that outrage improves user engagement, that people get easily attracted to hate speech, extremism and news that trigger jealousy, envy, anger and indignation. Thus, to make people remain engaged for longer periods, they only need to promote emotionally charged material over neutral information. Many of our news TV networks have also learnt these tricks to increase their TRPs. Algorithms were thus designed to display and highlight such contents and place them at the top of the users’ news feed, and to recommend them to users in a targeted manner.
By 2017, Facebook’s popularity in Myanmar was so overwhelming that it was almost synonymous with the internet. It was then that Facebook algorithms started promoting anti-Rohingya videos and materials propagating violence against them. As Yuval Noah Harari wrote in his book “Nexus”, 70 per cent of such video views came from Facebook’s autoplaying algorithms and 53 per cent of all videos watched in Myanmar were being auto-played for users by algorithms. Thus, instead of people choosing what to see, algorithms were choosing for them, which helped fan the flames of widespread anti-Rohingya sentiments resulting ultimately in the worst genocide in Myanmar, causing millions of Ro hingya refugees to flee the country. Facebook was the chief instrument for organising this genocide, without any accountability to anyone.
And not simply Facebook ~ there are numerous instances that by promoting emotionally charged, divisive and harmful digital contents that users find irresistible, the entire social media ~ YouTube, WhatsApp, Snap chat, Twitter, TikTok, etc. ~ have radicalised people, engaged them in lynching of innocents, hate crimes and racial attacks, and even influenced voters by targeting political ads. As the Nobel laureate economist Paul Romer said, “the digital advertising model pioneered by Google and Facebook poses a growing threat to democracy. These firms know more about citizens of the world’s democracies than the Stasi knew about East Germans.
They can exploit what they know without relying on the coercive power of a police state. They can adjust what people see and exercise control by a thousand nudges.” Their power can be gauged from the fact that in the USA, Facebook controls an estimated 59 per cent of all spending on digital political advertising, with Google controlling another 18 per cent. Not in vain Zuckerberg had boasted in 2019 that Facebook is the only candidate for the role of global arbiter for political discourse because its content systems are “more sophisticated than what a lot of governments have.” We live in a strange world where all our dearly held concepts about the nature and goals of power, facts, truth and justice have been turned on their heads. In this world, there is nothing called ‘fact’, it only depends on how you look at things.
To every ‘fact’ there is an ‘alternative fact’, to every ‘truth’, there is an ‘alternative truth’. Populism has now bec ome the dominant political philosophy in every democracy. As Harari says in Nexus, populism posits that there is no objective truth and that everyone has the right to discover and hold onto their own truths ~ it is only a question of whose facts and whose truth. Power is the only reality in such a world, and truth and justice serve only as the means to gain power. In a world that has lost such distinctions based on old societal values, one that cannot distinguish between what is good and what is bad for children, families, personal lives and for society, social media with its deep penetration into society and powered by AI that is no longer a tool but a powerful agent for human interactions, can upend the moral basis of our existence and undermine all the values we have held sacrosanct all our lives and tried to transmit to our children.
Unlike knives and guns that cannot decide by themselves who to kill, AI is the first technology in human history that can decide for us, shape and solidify our biases, and vitiate our existence with the help of a toxic social media by spreading incendiary misinformation at the speed of light, while giving governments unprecedented power to exercise surveillance over their people. When an industry gets dominated by a few players like social media, the standard economic tool is to encourage more competition, and so there are anti-trust initiatives seeking to break the tech giants like Google in USA into smaller entities. But for social media, even competition will be based on the same business model and hence cannot address the evils it propagates and promotes. The problem is not the social media, but their business model that is based on digital ads.
In their book “Power and Prestige”, the 2024 Nobel laureates in Economics, Daron Acemoglu and Simon Johnson have reflected on the role of social media in increasing inequality, de-democratising politics, radicalising public opinion, and increasing the controls exercised by tech czars of the ilk of Bill Gates, Mark Zuckerburg, Satya Nadela, Sundar Pichai, Jeff Bezos and Elon Musk, not only over national resources, but also of national and global voices. Social media over which these people hold a vice like stranglehold has now become a major problem for every country. Instead of living up to their promises of connecting people for better sharing of ideas and democratising human communication, social media has become an uncontrollable monster that governments all over are finding almost impossible to deal with.
AI has now become the most powerful instrument of controlling and amplifying baser human instincts, while tech czars keep assuring that these problems are short-term and rectifiable with more AI and that we need not be scared but move forward with technology while they continue to increase their profits. Their commercial success has neutralised every opposition from opinion makers: journalists, business leaders, politicians and academicians; they monopolise every discussion on policy regarding technology. “This is an oligarchy because it is a small group with a shared mind-set, monopolizing social power and disregarding its ruinous effects on the voiceless and the powerless. This group’s sway comes not from tanks and rockets but because it has access to the corridors of power and can influence public opinion.”
Since banning social media is not a viable option, Acemoglu and Johnson have suggested imposing a hefty tax on digital advertising revenue, to be paid not on the profits as is the usual commercial practice, but on digital ad revenues. It is too easy for these powerful multinational companies to hide profits by shifting their income or profits from high-tax to low-tax jurisdictions, hence the tax has to be on revenues earned. To put things in the proper perspective, the global digital advertising universe is estimated to generate revenues of around $600 billion in 2024 ~ 42 per cent of it accrues to Alphabet (Google’s parent company), 23 per cent to Meta (Facebook), and 9 per cent to Amazon.
For Meta, digital ads account for over 95 per cent of its global revenue; for Alphabet, around 77 per cent. The firms generate such colossal revenues by prioritizing automation, surveillance and addiction over people’s welfare and collecting and monetizing data by fuelling people’s anger and helping governments to disempower their citizens. It would be wishful thinking to expect responsible behaviour from them ~ blinded by wealth, they would stop at nothing to garner more profits. Acemoglu and Johnson proposed a flat tax of 50 per cent when their annual digital ad revenue exceeds $500 million, the high threshold to prevent unintended negative effects on new entrants in digital economy. The steep tax can push all forms of media/communication away from an ad based business model towards a subscription-based one, where revenues will be dependent on the quality of content and user experience.
This may then make the social media and search engines move away from digital advertisement towards more meaningful and useful content. AI and social media have infinite potential for transforming human society for the better, and society’s goal should be to encourage the development of technologies that help people rather than drive them towards anger and violence, for which social media must be disentangled from their current business model that relies on revenue from digital advertising. Only then they will be forced to reform themselves by moving away from the bad equilibrium in which they are presently locked.
The Nobel Laureates cite the example of smokingin the 1960s, 45 per cent of adult Americans were addicted to smoking when the government decided that it was too much of an evil, and though it took decades to push people away from smoking, it could be achieved through a combination of measures like taxation and awareness. Similarly, they hold that it is now time to impose a hefty tax on digital advertising to wean the social media companies away from their present perverted business model.
(The writer is a commentator, author and academic. Opinions expressed are personal)