As is clear to just about everyone looking back at past two years—let alone the past two decades—the digital world is rapidly changing beneath our feet. Since our entry into the 21st century, there has perhaps been no greater change than the growing dominance of technology over our lives. AI spearheads research, governments engage in cyberwarfare, and billions of people go throughout their day with powerful computers nestled in their pocket. As the previous essay mentioned, in this new landscape, data reigns supreme—bringing knowledge, power, and profit to whoever holds on to it.
The greatest weapon in this fight for data collection is undoubtedly social media, allowing unprecedented connection between humans, irrespective of geography. Social media urges users to share their lives online for all their friends (and Big Data) to see. It is the perfect data collection model, where users willingly post their hobbies, interests, and personal information, thereby doing Big Data’s work for them. While there are many social and societal problems with such a system, information has always been and always will be profitable, so it would be incorrect to point to the collection of data as social media’s core problem. Instead, the nature of the problem with social media is the way in which social media companies organize and categorize users in order to profit from them. By using user data to divide people into similarly minded groups online, social media companies are raking in profits and site engagement at the expense of dividing families, communities, and society.
Despite the changes brought on by social media today, using the internet to communicate with other people is nothing new. As suggested by its early name of “World Wide Web,” the internet was conceived as a means to bridge the geographical divide, to connect people, and to bring them closer together. However, this connection brought with it a deep sense of profound freedom, authority directly derived from consensus, and “the disappearance of the state” (Lessig 3). This was an online world in which regulation seemed distant, if not flat out impossible, whose “purpose was to enlighten, not monetize” (Snowden 44). Furthermore, in the early days the internet “was a place for conversation and exchange” more so than just checking email or stock prices (85).
Early Web 1.0 communities were usually chat forums formed around the discussion of a specific topic, such as computers, music, politics, text-based games, etc., and each community had their own set of rules. What almost every forum had in common, however, was anonymity. This structural choice provided two distinct societal advantages:
- Democratization of Input: Users facing discrimination in the real world—disabled individuals, women, minorities, and so on—now had a space “where [their] words were taken for what they meant,” instead of focusing on who said them (Lessig 87).
- Ideological Fluidity: Instead of having to “close our minds by picking sides” and doubling down to save face, one “could wake up every morning and pick a new face by which to be known to the world” (Snowden 47).
Even though millions of people had already formed gigantic social communities on the early internet, the communities one joined, the people one talked to, even who one was, were all a conscious choice. No meaningful metadata followed the user, and it seemed as if “the ‘Internet Button’ were actually a reset button for your whole life” (47). In the eyes of Edward Snowden and many others, this made all the difference.
Compare that to 2022, where privacy and anonymity are the exception rather than the norm. According to Lessig, the same code that once enshrined freedom may also “perfect control and make highly efficient regulation possible” (Lessig 4). This is precisely what has happened, and code today puts all the power in the hands of governments and corporations instead of users. Under laws such as the Communications Assistance for Law Enforcement Act (CALEA) and FCC edicts, internet architecture (code) was mandated to have built-in backdoors to make government wiretapping easier (Lessig 63-64). Furthermore, thanks to private-sector companies such as MaxMind and Geoselect, a user’s IP address can now reveal their location whenever they connect to the internet. Google searches—often tied not only to geographical regions but logged-in Gmail users—are considered by some to be “the most important dataset ever collected on the human psyche” (Stephens-Davidowitz 14).
The biggest change, however, has been the rise of social media in its modern sense, where users publicly go by their real names and connect with their real friends in cyberspace. The days of being able to simply reset one’s online identity at will are gone, and unlike before, the user on a site is now intimately linked with the person in real space sitting behind the screen. This, according to Snowden, is “the greatest iniquity in digital history” (46). While this is true, connecting one’s real-self to their virtual-self would be hard to avoid as society increasingly leaned on cyberspace for commerce, socializing, and providing government services. However, the direction social media companies have taken this development was completely avoidable and is not innately tied to the nature of cyberspace.
Social media is built on advertising revenues, and advertisers’ profits depend on user engagement, which is achieved using Big Data to provide users with content that keeps them scrolling. Through thousands of A/B tests and different ads thrown at users, a social media site such as Facebook “becomes a site optimized to maximize how much time people spend on Facebook” (Stephens-Davidowitz 220). We now arrive at the true nature of the problem with social media: as data and preferences are now tied not to the user on a site, but rather to the person themselves, individuals are now algorithmically isolated from critically engaging with differing perspectives in cyberspace.
In the eyes of many advertisers, “most people objected to advertisements because they were irrelevant,” and that they “would welcome pitches tailored to them” (O’Neil 69). This leads us to the logical conclusion that people will follow content they find agreeable, and walk right into an echo chamber, totally cut off from other sources of information. To an extent, this is true, as evidenced by the Cambridge Analytica scandal in which candidates like Ted Cruz “used [their] studies to develop television commercials targeted for different types of voters, placing them in programming they’d be most likely to watch” (191). Politicians also use Big Data to fuel confirmation bias, telling voters what they want to hear and making them “more likely to accept the information at face value because it confirms their previous beliefs” (187). The same principle holds true for social media companies and private-sector advertisers.
Interestingly enough—and in what may appear to be a contradiction to the echo-chamber argument—Seth Stephens-Davidowitz presents data showing that the internet is actually less segregated than real life. In his book, Stephens-Davidowitz shows that 45.2% of users on a news website hold opposing political views, compared to only 41.6%, 40.3%, or 34.7% of coworkers, neighbors, or friends, respectively (Stephens-Davidowitz 142). However, this does not mean opposite sides are engaging, as many of these users “visit sites of the opposite viewpoint, if only to get angry and upvote” (143). Yet, anger drives engagement, and “people can claim they’re furious…yet they’ll still click” (156). A click is success to advertisers, who will in turn feed more angering content to the user.
What this structural design optimization means for society is threefold:
- Bipolar Volatility: Users bounce between increasingly radicalized content which they agree with, and increasingly radicalized content that gets a rise out of them, strengthening the confirmation bias that their side is right, and the other is wrong—leaving the moderate middle ground empty and barren.
- Invisible Algorithmic Curation: While Stephens-Davidowitz asserts that people are more likely to be ‘friends’ on Facebook with political opposites due to the weak social ties on the site (144), Cathy O’Neil reminds us that Facebook has a history of tinkering with the news feed, meaning that certain content posted by friends—such as hard news—can be elevated or suppressed above the usual cat videos and photos (O’Neil 182-83).
- Enforced Conformity: With the lack of anonymity on social media, “you have a large incentive to make yourself look good,” which means hiding one’s true self to fit with others’ expectations of who you are (Stephens-Davidowitz 150-51). As a result, people are more prepared to dig in and defend their confirmation bias rather than admit they are wrong.
The result of this algorithmic isolation within social media has been the erosion of freedom, discourse, and choice in the online space. Users exist in their own social circles of confirmation bias, venturing outwards with the intent to be angry and argue, rarely with the goal of critically challenging and upholding their own views. The world through the lens of social media is rapidly resembling the closed corporate architectures of modern fiction—a closed ecosystem wary of, or even hostile to, anything that might upset that equilibrium. Unlike the early net, where joining each group or forum was a conscious decision by the user, a modern social media site like Facebook “determines, according to its own interests, what we see and learn on its social network” (O’Neil 180). With the most polarizing content garnering the most engagement, the middle ground where opposites once met is being covered by the rising sea of the far left and far right.
With the nature of the problem being clear, we must now turn to possible solutions. As was previously stated, social media is not inherently polarizing, rather the current state of the system is a conscious choice made by social media corporations and Big Data on behalf of the people. There is real good hidden in the noise of social media—friendships, learning, potential for discourse, solidarity. The same sense of community Edward Snowden found on BBSes (2000-word answers to questions, offers of free computer parts, encouragement, etc.) still exists today, but is so much harder to find. Seth Stephens-Davidowitz sees good, too, in the form of a ‘digital truth serum’ such as “knowing that you are not alone in your insecurities and embarrassing behavior,” as evidenced by raw Google search analytics (Stephens-Davidowitz 158).
Lessig reminds us that code is law, that everything we see online is by design, with values built into the space that control our interactions with it (Lessig 6). The current design of social media prioritizes polarization and ideological isolation to drive advertising profits. It does not have to be this way. We should heed Lessig’s argument that legislation can be used as “a tool to change incentives” behind online architecture (262), and that collective action is the way this will be achieved (232). Perhaps a rethought Fairness Doctrine for social media could be the answer. Maybe AI could help fight polarization and disinformation. The specifics are not yet important—but the commitment to debate and public discourse about social media’s structural role, is.
As we continue our path through this century, social media’s influence in our lives will only continue to increase. The COVID-19 pandemic forced the entirety of society online, giving us a clear view of both the necessity and danger of social cyberspace to our well-being and our democracy. What we have lost online is the ability for widespread discourse, freedom of expression, and choice, as the current profit-model for social media corporations requires feeding users polarizing content to spur site engagement. So far, we have sat idly by, watching social media’s potential for good being destroyed by the invisible hand of advertising profits. This is entirely within our control; we only “make the hand invisible by looking the other way” (Lessig 339).
The solution is legislation directing the architecture of social media—but laws do not come out of nowhere. Acknowledging the problem is the first step. There is a growing sense of unease in society about social media’s role in our lives. This sentiment must be harnessed and turned into deliberate structural changes. Only then may the core problem with social media be solved.