May 5 has come and gone, and former U.S. President Donald Trump is still locked out of Facebook. The decision by the world’s largest social media company’s advisory board has left no one satisfied. The right wing is outraged that Trump’s account was not reinstated; the left wing is outraged that Facebook didn’t ban him permanently from the platform. Instead, the advisory board merely postponed a final decision, setting a six-month timetable for Facebook management to come up with a more robust rationale for any ban.
The decision papered over more serious questions than many in Silicon Valley can yet answer. Namely, can social media companies like Facebook and Twitter reduce or reverse the socially destructive impact of their platforms—specifically, the tendency of their algorithms to stoke divisions and conflicts? More importantly, can these companies create business models that replace a destructive attention economy that relies on pawning detailed user data and is so easily gamed and abused by bad actors?
The answer to this question may come from an unlikely source: Apple. The company has anointed itself the privacy gatekeeper for its legions of iPhone customers. Apple is now threatening to choke off the flow of data that feeds algorithms used by social media companies to target ads and content and enable the abuses.
The primary problem with social media isn’t that it has provided a platform for white supremacists, conspiracy theorists, violent extremist groups, or populist politicians with a history of spreading lies and inciting violence. It is that such platforms are designed, at their very heart, to multiply the messages of these groups to users who might be most swayed by them.
The matching engines at Facebook don’t care if they get more clicks by pushing posts denying the Holocaust; Facebook’s algorithms are soulless and care only about clicks as they translate into money. The company’s content moderation is a fig leaf and easily abused by groups of users reporting accounts they don’t like. Social media’s algorithmic echo chamber automatically and inexorably amplifies much of the most dangerous, salacious, offensive, and generally destructive content. If Apple cuts off the data that drive algorithms, these machines will flounder and sputter.
The company, built by Steve Jobs, has no skin in the social media game and is putting its users’ privacy before the greed of application developers. Apple isn’t doing this out of the goodness of its heart, of course. It makes money by selling hardware, warranties, and subscription fees associated with iCloud and Apple Music; indeed, walling off its platform to control monetization has long been part of its business model.
It’s not surprising this puts Apple at odds with Facebook. In late April, Apple enacted a policy that all applications on its iPhone platform must ask permission before they can collect user data. This will force social media companies to develop business models that are less reliant on targeting. Facebook will be especially affected because of the granular targeting it offers, not only on its platform but also throughout its advertising network.
The early returns from the release of Apple’s update show that 88 percent of worldwide users—and a stunning 96 percent of U.S. users—refused to allow apps to track them. It is doubtful whether Facebook and other social media platforms can sustain their current business models when such a large group of users are blocking access to their data. Indeed, collecting user data advertisers can use to finely target ads is the core business model of the social media giants. Facebook now relies on ads via its mobile applications for the majority of its revenue. The iPhone is a crucial part of that revenue stream.
This is why Facebook has strenuously objected to the policy change, saying Apple’s move would harm small businesses that market on the platform. Left unsaid was who else might be harmed: Trump’s campaign. Though both major parties have mounted vigorous efforts to use Facebook for fundraising, Trump has been a pioneer in building his campaign largely on an extremely sophisticated Facebook advertising effort to collect donations, email addresses, and mobile numbers. In fact, Facebook employees were ensconced in Trump’s campaign as early as his 2016 presidential run, playing a key role in Silicon Valley companies’ technical and advisory support for Trump.
Depriving the algorithms of data may go a long way toward reducing the worst impacts of social media. We don’t know the long-term outcome, but Apple is imposing a fascinating social media experiment on Facebook, Twitter, and other social networks that may yield some surprising results. Perhaps this experiment will force the networks to search for better ways to make money than by vacuuming up user data and behavior to feed targeted ads and angry mobs.
Evidence already exists to challenge the assumption that social media can work only with targeted ads. In 2019, to protect itself from potential fines under the European Union’s General Data Protection Regulation, the New York Times cut off all behaviorally targeted advertising and replaced it with contextual ads (that is, ads relating to the content of the page rather than the user) and geographically relevant advertising. The paper’s ad revenues did not plummet. On the contrary, digital ad revenues increased significantly after these steps were taken. In all likelihood, the newspaper’s readers appreciated not seeing ads that were obviously tracking their personal data.
Dismantling the worst elements of the attention economy will be a challenge. Thousands of players are vested in perpetuating this model, including influential politicians and their campaigns, giant advertising technology companies, and businesses that believe they can reach users only through targeted ads. But the attention economy is the only internet business model the world has ever really known. The die was cast in the early days of online advertising, and no alternatives were ever seriously considered once the Google juggernaut got underway. It may turn out that for Facebook, Apple’s new emphasis on user privacy is the proverbial gift horse. By forcing Facebook to consider other business models, Apple may force it to consider how to drive healthier forms of engagement.
In depriving Facebook of behavioral data, Apple will stop giant pattern-matching algorithms and blunt both the impact of and backlash against social media. None of this, however, implies Facebook can avoid having to make a decision on banning Trump. Using social media platforms to incite violence, such as the Jan. 6 insurrection, still crosses a line. That said, there is no conceivable way to police all the content flowing through the platforms. And the very algorithms so expert at matching users primed for anger with memes that make them angrier still are helpless in judging the content and context of posts. The final decisions take human judgement.
That’s why Trump is a symptom rather than a source of the problem. And it’s the deeper illness that must be treated. Taking painkillers is an obvious response to a painful tooth, but removing the tooth or stopping the infection that caused the pain is the only way to feel better in the long run. There will always be figures like Trump riling up unsavory groups on social media. But defanging the algorithms and creating a different business model will reduce a problem that’s presently virulent and out of control to one that’s manageable.
In the end, Facebook may thank Apple for its inadvertent assistance. And all of us may end up with social media platforms that are no longer personalized echo chambers designed to profit by stoking humanity’s very worst impulses.
