Most of us have probably seen The Social Network, or at least heard about how it stages Facebook’s primal scene. One night, a bored Mark Zuckerberg (Jesse Eisenberg) does some hacking and coding — dramatized, in Hollywood style, as his ability to type really fast — to set up a website that lets visitors judge the hotness of the women of Harvard. It proves so popular that it threatens to bring down Harvard’s entire computer network. Here was the kernel of Facebook, with a foretaste of its worldwide success.

Aaron Sorkin’s screenplay cleverly frames this mini-narrative within a broader story about social anxieties. In his fictionalized account, we first meet Zuckerberg as he ruins an otherwise promising first date by ranting constantly about his own questionable social standing at Harvard. And in the end, after he has bested the photogenic and entitled Winklevoss twins (as well as one of his more naive friends) in the legal struggle over the rights to the Facebook concept, we see him hitting refresh continually as he waits for that same long-lost first date to respond to his friend request. Sorkin invites us to conclude that Zuckerberg built the “hot or not” machine with the long-term goal of getting voted “hot” himself — or, perhaps more ironically, that he has been taken in by the same manufactured social anxieties that he is now profiting from.

Though it is not a strictly factual account of Facebook’s origin story, The Social Network gets the platform’s animating ideology right. While it has evolved into something far more complex than its “hot or not” roots, Facebook is still a technology for passing judgment. It may not offer, as Reddit does, an explicit way of expressing direct disapproval, but every status update and comment is still staged as a mini “hot or not” contest: Either you press “like” or pass over it in silence.

Every status update and comment is still staged as a mini “hot or not” contest: Either you “like” or pass over it in silence

Recognizing that “like” is not always an appropriate reaction, Facebook diversified the function to allow users a range of judgments corresponding to the range of emotions we learn to name in kindergarten (love, laughter, surprise, sadness, and anger, alongside the classic “like”). Still, the core gesture is the same. Even the negative emotional responses are meant to echo, and thereby express approval of, the negative emotions expressed through the post itself. The judgment is slightly more nuanced, but at bottom it is still a question of “hot or not.”

It turns out that making Facebook do anything but “hot or not” is hard. For instance, some people try to make it into a platform for sharing interesting links. But in practice, that effort devolves into the activity of passing judgment on those links — and more important, inviting others to pass that same judgment and thereby express approval of us. And in fact, this system of judgment is impossible to opt out of, because Facebook does not allow you to turn off comments or the numerical displays of “likes” and other emotional reactions. Facebook makes us all into the fictionalized Zuckerberg, awaiting a positive judgment from the chain of judgments we set off — or, more precisely, a positive assessment in the form of that chain. What is most “hot” turns out to be what is most viral. We measure our status by how far our internet-transmitted conditions have propagated themselves.

This leads to the much-discussed phenomenon of passing along links without reading them or checking their source. For Facebook users, the important thing isn’t typically to spread accurate information — after all, people in a given social circle often share the same links every day — but to elicit a reflection of our own correctness, whether it is expressed with smiley faces or frowny faces. We can tell we are correct by the fact that everyone approves of us, which is to say that everyone has caught the same virus. Self-expression and conformity strangely overlap as we compete to see who can be the most alike, meaning who can be the first to say “what we’re all thinking.”

Attempting to use Facebook as a discussion forum may seem to have better chances. Alongside the infrastructure of “liking,” we initially appear to have nothing but an empty box — a free-form discursive medium — that, when filled in, gives birth to a comment thread not unlike those found in nearly every internet discussion tool since Usenet. Here, too, however, the inertial pull of passing judgment is strong. Productive discussion requires at least a modicum of critical distance, a willingness to entertain unfamiliar or even opposed views for the sake of argument, but the Facebook interface throws up obstacles to this. Inscribed in the comment box itself is a little picture of you, making every comment personal — as much about you as about what is being said. Indeed, as you scroll down the page, you see your own picture over and over again, inviting you to propagate your image further and further. And when others appear in the comments to your status, the same dynamics apply as with shared links — they are an opportunity to pile on, either with praise or with abuse, reinforcing the mutual regard of the “hot.”

Twitter lacks Facebook’s genealogical roots in a literal “hot or not” contest, but it too is a technology for judgment. Users seek followers, retweets, and (as a kind of consolation prize, since they do not present themselves so forcefully) likes. In many ways, the dynamics are the same as Facebook, only faster, because of the rapid-fire nature of the shorter format and the unfiltered structure that, unlike Facebook, lets you see everything your friends post immediately. This makes Twitter much harsher as well.

As a longtime Twitter user, I have often signed on after even a short absence to find my entire timeline consumed with outrage about an issue no one had been discussing previously. An instantaneous orthodoxy springs up, as when in early January everyone on left-liberal Twitter seemed to spontaneously decide that reimportation of prescription drugs was a non-negotiable priority and Cory Booker was a virtual demon for voting against it. This is not to attack the policy or defend Booker but to point out the strangeness that a relatively obscure point of policy, one that has been embraced at times by both Democrats and Republicans and appeared in Trump’s platform, should suddenly become a left-liberal shibboleth.

Ideological bubbles or fake news or harassment campaigns are features. A good-faith reader is worth one click, a devoted racist troll could be worth dozens

This kind of insta-groupthink is made much easier by Twitter’s design. If my previous paragraph were translated into tweets, my disclaimers could have ended up separated from my expression of skepticism about the importance of drug reimportation — leading to me being attacked as a Bookerite neoliberal shill whose very existence is Why Trump Won. Although threading tweets may mitigate this tendency, in my experience, people tend to the full thread only slightly more often than they read beyond a long article’s headline. The inconsistently observed custom of numbering tweets in a rant does little to push back against the inertia of decontextualized instant reaction.

The true danger on Twitter, though, comes when an ironic remark is deliberately decontextualized and circulated among audiences beyond one’s inner circle, where everyone is in on the joke. This can expose one to an almost unimaginable wave of vicious attacks, as when a black woman academic was singled out by right-wing commentators for suggesting that white men sometimes cause problems in her class. This observation was taken as a blanket judgment of all white men, a case of anti-white racism, etc., etc., and thousands turned up in her Twitter responses (as well as in her email inbox and those of her employer).

In cases like these, Twitter becomes less a popularity contest than a reality show, where disfavored users are “voted off” through systematic harassment that essentially renders it impossible to use the platform any longer. The harassers are in a sense gaming the system, but the system was already a game of accumulating approval and expanding your network. The very same design features that demand that we continually increase our “exposure” (in the contemporary, positive sense) also leave us exposed in the more conventional sense of being vulnerable and unprotected. And even if the designers of this site did not consciously intend this outcome, it is already implicit in the way that their systems prey on our vulnerability to social anxiety, our exposure to social approval — and disapproval.

Whatever else the internet is, then, the hegemony of these forms of social media have turned it into an increasingly efficient machine for judgment — for passing judgment, for eliciting judgment, for soliciting judgment. And the more attention you attract, the more likely it becomes that you will face an overwhelming backlash of negative judgment. Andy Warhol famously asserted that in the future, everyone would be famous for 15 minutes, but he didn’t specify what form that fame might take. Were he still alive, he might now say that in the future, everyone will be hated by thousands of strangers for 15 minutes.

Why are we so addicted to judgment, both dishing it out and receiving it? Perhaps that question can be answered by reframing it: Why do we turn away from — or, in many cases, preemptively reject — dialogue? In one sense, this may seem like a false dichotomy. After all, what is agreement or disagreement, or even the simple weighing of an argument, if not a judgment? Indeed, Kant goes so far as to characterize every mental act as one of judgment. What is new and distinctive about social media judgments is, first, that they foreground and render explicit what was previously implicit — above all by quantifying approval. Second, and more insidiously, they direct the force of judgment away from the “content” of the exchange and toward its participants.

By means of this amplification and redirection of our faculty of judgment, the forms of social media built on metrics tend to breed certitude and spite rather than the critical distance from one’s own views and a willingness to entertain those of others that dialogue requires. But it’s not just a question of our having picked up bad habits from social media. The truth is that dialogue is risky, because your efforts may not be rewarded with new insight. Indeed, you may be played for a fool by a bad-faith interlocutor who is purposefully trying to waste your time or even elicit condemnable statements from you.

We measure status by how far our internet-transmitted conditions have propagated themselves. Self-expression and conformity overlap as we compete to say “what we’re all thinking”

Social media can be the worst of both worlds in this regard. We are brought more and more into contact with total strangers, many of them completely anonymous, meaning that the element of trust is totally lacking, and at the same time, the easily trackable, archived nature of most online discourse means that any of us could be sitting on a ticking time bomb just waiting for a sufficiently dedicated troll to set it off, by propagating it in the public field of judgment.

And what motivates them to do this is that judgment, unlike dialogue, has an immediate, guaranteed payoff. You get your pellet of self-satisfaction every time you hit the lever. Not only does passing judgment let us enjoy a feeling of strength and rightness individually, but we often pass judgment as part of a larger group — meaning we get to experience approval and solidarity as well as a license for cruelty. At least for a moment, that feeling of belonging and that cathartic experience of venting frustration and resentment can help to distract from the harsh realities of a world where economic precarity leaves people feeling less and less in control over their own lives.

At this point, it would be easy to slip into the familiar routine of scolding the laziness of internet users nowadays — casting judgment yet again! I think that would be a mistake, however. The problem is not only that many people are too lazy to read carefully, though surely they are, at least sometimes. Nor is it that people are too impatient to engage in genuine dialogue, though again, they often are. I would even go so far as to say that the problem is not simply, or at least not directly, that people are uneducated or willfully educated. All these factors are real, but they are symptoms rather than causes.

We could think of judgment as the fast food of the internet, and judgment-driven social media as fast-food restaurants. The problem isn’t that a critical mass of individuals are choosing to eat fast food instead of going to the farmer’s market and making time to cook at home. The marketplace of the internet is not simply responding to some kind of intrinsic desire for online fast food on the part of individual consumers. Rather, it’s that the ad-driven, click-driven, gamified business model that drives the culture of judgment makes fast food the only economically viable option. In short, this economic model is turning the internet into a food desert.

The internet as we know it today is an overwhelmingly ad-driven enterprise, and that means that it is click-driven. Where print and television advertisers had to rest content with only indirect evidence that their message was reaching customers, the internet provides the opportunity to measure user engagement directly. And when platforms can generate measurable user engagement through content the users create and spread themselves, why would they invest in anything more? Why pay writers when a Twitter user can reach thousands? And what more efficient way to drive ad revenue than by “rewarding” users according to metrics that directly correlate with engagement and therefore ad visibility? By turning users into relentless self-promoters, you also necessarily turn them into promoters of the site that enables their self-promotion.

These material incentives line up to make the most toxic aspects of contemporary internet culture — things like ideological bubbles or fake news or harassment campaigns — unfixable under the most prevalent media business models. From the perspective of driving user engagement, these behaviors are features. This is why newspapers allow readers to vandalize articles with racist and otherwise hateful comments — engagement is engagement! A good-faith reader is worth one click, whereas a devoted racist troll could be worth dozens.

Twitter is reluctant to ban trolls for similar reasons — they increase engagement from friends and enemies alike, building group solidarity through shared enmity. It’s also why harassment campaigns are probably a permanent feature of public-facing social media. For every user who gets “voted off” (i.e., driven from the public sphere organized by the platform), hundreds and perhaps even thousands enjoy the adrenaline rush of the mob violence.

I’m not sure how to solve the problem, but I am pretty confident that the answer isn’t to whip up even more online outrage. Scolding friends who share so-called fake news may have its place, but when people in food deserts “choose” fast food, the problem isn’t their lack of willpower. It’s the paucity of choices in the first place. Just as with a real food desert, it would be a mistake to treat the social media food desert as a place where better options just happen to be less profitable. The active anti-intellectualism of the internet is part of a broader hollowing-out of institutions of education and information, which began before the internet became a mass phenomenon but which the click-driven internet has helped accelerate. Just as the social media phenomenon is driven by a business model, so too is the broader deinvestment in education part of the neoliberal economic model that privileges short-term profit over long-term investment, or more broadly, the accumulation of wealth over any other social goal.

There are some possible paths forward. I am encouraged, for instance, by the development of nonprofit, open-source social media platforms like Mastodon, though I worry that their fully decentralized structure would make it difficult to take concerted efforts to prevent systematic harassment from taking root. A retreat to older forms like blogs might also help us cultivate spaces for dialogue rather than judgment. While the blogosphere was not a utopia of enlightened discourse, it did at least open up the possibility of building a long-term community centered on ideas and discussions, whereas Facebook tends always and everywhere to turn even the most enlightened and informed conversation into a popularity contest by proxy. Yet reviving blog-based discussion that may be an uphill battle. In my experience with blogging, I’ve found that it is nearly impossible to coax a critical mass of readers to comment on the blog post itself rather than sharing the link on Facebook and discussing it there.

But the problem is bigger than the internet, and so the solution must be as well. We must restore institutions of education and information and rebuild the decaying social solidarities for which the social media culture of judgment serves as a hollow substitute. That is to say, the problem is not ultimately technical, but political. Better online tools may help us do that work, but they can’t do it for us.