To fix Facebook, we need to socialise the networks

Last month, former Facebook employee Frances Haugen brought documents to the press, giving further evidence about what is already known — that Facebook prioritises profits over human rights. Quality of information, humane standards and the well-being of teenage girls on Instagram are sacrificed when in conflict with maximising profits.

For all the good that Haugen did by bringing new information to the public, she also framed the issue in terms sympathetic to Facebook. Haugen joined Facebook in 2019 — when it was well-known the company was awful — on the premise that “Facebook has the potential to bring out the best in us.” In her testimony to the US Congress, she continued with this sentiment, stating that Facebook needs to be made better rather than replaced altogether.

What mainstream commentators won’t say is that Big Social Media cannot be fixed within a capitalist framework. Moreover, US intellectuals won’t address the fact that Facebook is a crown jewel in the American empire project, a profit-seeking corporation that colonises foreign markets. If we want a solution to Facebook — and the rest of Big Social Media, including surveillance-based corporations like Twitter, YouTube, and TikTok — we need to socialise the networks based on a digital democratic commons. For this, we already have functioning prototypes like the Fediverse in place.

The Fediverse: Decentralised social media networking

As I detail in a long-form article, Social Media Socialism, from the time that Big Social Media was hatched in the mid-2000s, hacktivists in the Free and Open Source Software community began developing an alternative designed to liberate users from fenced-in corporate networks like MySpace and Facebook. The Free Software community sought to decentralise social media technology so that people would no longer be locked into corporate silos — who then determine the speech codes, rules, and behaviors that everyone must abide by — so that the Big Social Media corporations can accumulate wealth for their rich owners and executives.

Big Social Media networks can retain their users in part because they do not allow their networks to interoperate. If you’re a Facebook user, for example, you cannot befriend or follow a Twitter user, and vice-versa. You cannot “like” or add comments beneath the post of another network. As a result, if you want to participate in Facebook or Twitter, you have to create a separate account for each, and your interactions are entirely contained within each respective network. This dynamic means for every user that joins your network, the network becomes more valuable, a phenomenon called “network effects”.

If you want to join a new network, you have to convince all your friends to come with you, and they have to convince all their friends to join them as well. This dynamic concentrates social networks because people are not willing to do this often: it is only practical to join a few networks, as nobody wants to log into fifty different social networks, each functioning as if on a separate island.

Decentralised social networks were developed to fully interoperate and prevent this kind of vendor lock-in. After a decade of development, social media decentralisation finally caught on with the rise of Mastodon.

Mastodon is a Twitter-like social network that allows users to create their own Mastodon sub-networks called “instances”. Users can also interact with other users of social media networks external to Mastodon itself, in what’s called the Fediverse, so long as those networks are using the same “protocol” to talk to each other. (The protocol used is essentially a common set of standards for interactions across social networks.)

In time, a handful of new Fediverse networks were built, including PeerTube (a YouTube clone), Pleroma (a Twitter clone), and PixelFed (an Instagram clone). To date, Mastodon is the most popular Fediverse network with over two million registered users.

Mastodon offers users the ability to join or form their own Mastodon instances so they can self-manage their user data and rules. A Mastodon instance might only allow cat pictures, for example, or it might be dedicated to sports, politics, or general-purpose use. Users are given names that look like email addresses: e.g. you might be “@[email protected]”. You can then interact with others on the mastodon.social instance, but also with people at other instances, just as you can email people across mails servers (Gmail, ProtonMail, Yahoo and so forth).

Each instance can decide if it wants to interoperate with other instances. If an instance is not desirable to some people — for example, let’s say it is loaded with extremist content — it can be banned by network administrators. Individual members can also choose to ban other instances for themselves, so they don’t have to see content posted by that network. When a right-wing social media network Gab was shut down and its users migrated to Mastodon, many Mastodon users banned the Gab instance.

Mastodon has three time lines instead of just one. The first is “home”, which displays posts from the users they follow. The second is “local”, which displays posts from the network they join (for example, mastodon.dogs). The third is the federated time line, which displays posts from all users their network federates with. Users can display all three time lines on their screen at once, or they can choose to display just one at a time.

The Mastodon platform is also Free and Open Source Software, which means it can be customised by other software developers who might want it to look or function differently. Librem Social, for example, took the Mastodon code and altered it so that there are no timelines — you only see posts from people you follow.

Networks like Mastodon and Librem Social could build in algorithmic filtering, but at present they do not. Algorithmic filtering has often been used by Big Social Media networks to maximise user engagement, so that users see more ads and make money for the platform owner. Fediverse platforms do not display ads and do not seek profit, so they do not engage in user exploitation to maximise profits.

Content moderation is left to network administrators and individual users. Many Fediverse networks require users to label graphic or controversial posts as “not safe for work” (NSFW) and remain hidden unless you click “display”. As one can expect, with so many users able to instantaneously communicate with each other, content moderation remains a challenge. The best that can be done to reduce social harms in large networks is probably to have humans hired to manage it. This could be covered by the public purse, but is quite a costly endeavor given the sheer volume of social media traffic. It’s a problem afflicting all social networks.

While Fediverse offers privacy from giant centralised actors like Facebook and Twitter, users can still be surveilled by network administrators. This is because the technology is built to channel user activity through each network server (instance), which hosts the data and creates a record of user behavior. Another product called LibreSocial is attempting to improve on this feature by having users host the data and transmission on their own devices, with data shared in a peer-to-peer fashion. Data not meant for public sharing is locked away behind encryption; only the intended recipient can unlock private content.

In time, decentralised social media could form a network of peer-to-peer social networks.

Social media commons vs capitalist social media

The Fediverse offers a real-life example of how social media can be run as an ad-free, nonprofit, community-controlled system. One of its most important features is that it undermines the corporate, profit-seeking, American-dominated world of social media. This is critical, because the world’s people — not just Americans — are subjected to the power of Big Social Media.

Unfortunately, American intellectuals in the mainstream ignore the Fediverse and the option to fully socialise social media as a commons. Instead, these critics predominantly favor using antitrust to break up Facebook into three parts: the Facebook network, Instagram, and WhatsApp ( leaving networks like Twitter and TikTok alone). 

They also favour a limited form of interoperability that forces only the largest social networks to allow others to interoperate. This very partial form of interoperability will likely leave Big Social Media at an advantage, as small networks remain incentivised to interoperate with them so that their users can interact with members of the giants, but less so with the other small social networks they compete against for market share.

For antitrust law, competitive capitalism is the ideal state of an economy. Yet there are strong reasons to doubt that competitive capitalism will fix social media networks. To the extent that social media is privatised, it will remain problematic because the same exploitative dynamics persist: in order to maximise revenue, profits, growth, and market share, a network must maximise user head count and time spent on the network. The capitalist war of competition for eyeballs is a problem, not a solution.

The notion of force-feeding people ads, which inflames consumerism that is destroying the planet, is not addressed by antitrust advocates. Similarly, antitrust advocates do not explain why social media networks should be owned by corporations, which place profits over the public interest and concentrate wealth into the hands of executives and shareholders. And in the US, they do not address digital colonialism by which American tech giants like Facebook, Twitter, and Google prey on the Global South.

A solution which fully socialises the networks would best serve the global public interest. To accomplish this, Big Social Media networks can be given a grace period before they are open-sourced, converted to public property, and forced to decentralise. Citizens could be given tax-based vouchers and use them to fund networks and content moderation services of their own choice. Researchers at universities and publicly-funded technology could support the development and maintenance of networking technology. In South Africa, universities and organisations like the Council for Scientific and Industrial Research could devote resources to developing and maintaining social media infrastructure. User access would be equitable and ad-free, while communities would have a real say in how their networks operate.

It’s time for the activists and intellectuals across the world to consider commons-based solutions like the Fediverse and LibreSocial instead of capitalism as the organising system of social media. A different world is possible, but it will take a fundamentally different approach to get us there.