The Perils to Democracy Posed by Big Tech
Silicon Valley has changed. The same companies that once held free speech so sacred they refused to ban terrorists and stood idly by as their tools were used for genocide have suddenly decided that censorship is the only way to save the web in the Era of Trump. The platforms that control what more than a quarter of the earth’s population say and see online have pivoted from the “free speech wing of the free speech party” to an Orwellian world in which unpopular speech can simply vanish. What do these changes portend for the future of free expression and democracy itself?
The election of Donald Trump was a watershed moment for the nation’s technology and media companies. Almost overnight, phrases like “fake news” and “filter bubbles” became household terms as the nation’s elite attempted to explain an election they could not understand. Silicon Valley’s response was to embrace censorship, viewing aggressive control over the online world as the best way to tame a changing electorate. The news media responded by claiming ownership over “truth” and hardening its coverage of the president from dismissive to adversarial.
Freedom of expression has long been considered the bedrock of American-style democracy, ensuring the free flow of ideas, no matter how unpopular they may be. Although it was always easier for the wealthy and the intellectual elite to broadcast their views to the masses, this right was considered sacrosanct. Social media has upended the natural order, however, and as elites lost their ability to control the conversation, free speech has been recast from the pillar that supports democracy into the gravest danger that threatens it.
Social platforms helped Black Lives Matter raise awareness and allowed #MeToo activists to tell their stories. Yet, as a rising tide lifts all boats, so too has social media given voice to the toxic fringes of society. The anonymity and global reach of the platforms have helped spread everything from conspiracy theories to almost unimaginable racist, sexist, and dehumanizing speech.
The result is an online world that has become so filled with hatred and threats of violence that it increasingly can no longer serve its role in facilitating the exchange of ideas necessary to a democracy.
How then to salvage the ideal of free speech amid these competing and conflicting forces?
This question was the focus of Google’s “Good Censor” briefing that was leaked last month. While many commentators seized on the document as a de facto roadmap to conservative censorship, the reality could not be further from the truth.
I should know, I was one of the three external experts interviewed in its creation (though I was not aware at the time that it was for Google). Rather than a roadmap to censorship, Google’s briefing represents precisely the opposite. In short, it focused on how to ensure that the most vocal and toxic fringes screaming blindly on all sides do not drown out the myriad moderate voices in the middle having conversations.
I have long argued that the solution to the web’s toxicity is to nudge society away from chasing viral fame and toward having real evidence-based conversations about our shared future. Social platforms have instead chosen to craft secret lists of what is “acceptable speech” and employ thousands of human “moderators” and black box AI algorithms to silently scour their platforms and delete anything that falls afoul of those rules.
Such an approach is problematic when the very concept of what constitutes “hate speech” is fiercely contested. Few would disagree with banning a neo-Nazi who calls for violent assaults against synagogues. On the other hand, would a tweet by a professor endorsing the Boycott, Divestment and Sanctions (BDS) movement be banned as anti-Semitic hate speech? To the academics advising social media sites on their “healthy conversation” policies, the answer is almost certainly no. To many Americans who’ve looked at the ugly anti-Jewish statements of various BDS students on college campuses, the answer is an unequivocal yes.
As free speech has increasingly become an ideological battleground, the traditional differences between conservatives and liberals have been recast. Today, being a Republican is portrayed in some quarters as equivalent to Nazi membership. Driving this is the increasingly popular narrative that all of the fringe voices of hate online stand for the Republican Party in a way that those on the left do not stand for the Democratic Party. A Trump-loving mail bomber is held up as a direct product of the GOP, while a Sanders-volunteering shooter is dismissed as a disaffected loser with mental health issues. A Trump-hating mass killer is rewritten into a Trump disciple.
Silicon Valley has become increasingly enamored of the idea that all hate online is conservative speech. After all, it is far easier to just build a filter that blocks all conservative news outlets than it is to actually combat hate and misinformation. Unfortunately, in doing so, the social platforms have transformed a bipartisan fight against hatred into an ideological battle against our societal differences.
The employees of social media companies skew heavily liberal and thus it is no surprise that the moderation guidelines that govern what is permissible to say or see within their digital borders mirror those sensibilities. The companies themselves staunchly defend their practices, arguing they are ideologically neutral. Yet, when asked to support their claims with hard numbers that prove they do not remove conservative speech more often than liberal speech, they refuse.
The companies do not set out to preference liberal ideals to conservative ones. It is merely the inevitable byproduct of implicit bias in which a largely liberal workforce builds systems that mirror its own perspectives.
Newsrooms across the country also have unwittingly embraced this bias. When the president criticizes biased coverage, the reaction is one of dismissal rather than reflection. When CNN and MSNBC barely mention Hillary Clinton’s emails, their silence is viewed not as editorial bias but rather as differentiating between “reality” and “conspiracy theory.”
Unfortunately, these very real issues -- so corrosive to the functioning of democracy -- are lost in political hyperbole and emotion.
All the while, the technology companies that increasingly wield near absolute control over the online world decide what we are permitted to talk about. In turn, they delegate to the similarly opaque world of fact checkers the determination of what is “truth” and what is banned “fake news.”
Perhaps most dangerous of all, Facebook has increasingly begun to leverage its power to directly interfere in the voting process itself. Across the world, the company now sends get-out-the-vote reminders that it has shown to have a significant impact on voting behavior, yet it is unable to explain why not all voters see the messages or prove that the reminders do not result in a partisan skew.
The potential consequences of all these related concerns cannot be underestimated, or be left unaddressed. In the end, free speech is too important to democracy to leave to the shadows.