Redesigning Twitter for a network, not a broadcast

Twitter is a major social media platform that needs no introduction. With a mission statement of serving the public conversation, its values statement interprets Twitter as analogous to a public forum. Everyone can state their opinions, and everyone can be heard. As of recent, Twitter’s acquisition by Elon Musk has spearheaded controversial decisions to double down on the ‘town square’ free-speech principles of the platform, sometimes to the detriment of the balancing interests of user protection and public safety. As someone interested in understanding how design decisions shape how we interact with other people, I’d like to explore how alternative value designs could affect these interactions and the tradeoffs of these decisions.

The guiding principle behind this value redesign is to place less emphasis on Twitter as social ‘media’ and foster design decisions to place emphasis on Twitter as social ‘network’. I draw a distinction between the ideas of social media and social network to highlight the connotations of each. Media is news, entertainment, politics; it emphasizes the broadcast element of one-to-many communication for spreading ideas and influencing people. Networks are the connections between users and the people they engage with; they emphasize relationships and interactions within a community.

To that end, we can use Lawrence Lessig’s pathetic dot theory¹ of how artifacts become constrained to interrogate how Twitter’s current design promotes the ‘media’ in social media. I’ll primarily focus on the architecture constraint, while briefly discussing how architecture decisions can influence norms, markets, and law as well. Furthermore, I’ll be using the ‘Human Impact’ segment of the Security Cards brainstorming toolkit² (developed by Denning et al. at the University of Washington) to consider how these constraints can be exploited by adversaries and other bad actors. Reconsidering the values behind Twitter’s architecture is likely the most actionable step, as Twitter has the most agency over its own engineering.

Existing architectural decisions by Twitter unsurprisingly influence the interactions by its users. By default, Twitter accounts are public – anyone can see anything that any public account has posted. The retweet button enables the one-to-many style of broadcast communication for messages to be forwarded with potential exponential reach. Quote-retweets afford an additional layer of commentary on the message being shared. Post metrics like comment, like, retweet, view count, and account metrics like follower count, create incentives to pursue engagement metrics. Users with blue checkmarks get priority and better visibility over non-checkmarked accounts.

Architecture’s Impact on Norms

The one-to-many broadcast communication style draws in creators and journalists who use Twitter to communicate with their audience. The norms surrounding the blue checkmark are seen as a status symbol by some circles, and as a distasteful “selling out” by other circles. Discourse culture on Twitter often involves reactionary responses to controversial tweets, with metrics such as the like or quote-retweet ratio used to signify winning or losing arguments.

Architecture’s Impact on Law

While not exclusive to Twitter, recent moral panics about screen time, data privacy, and algorithmically generated news feeds have made their way to Congressional hearings with the threat of stricter regulation from public officials. One-to-many communication, where each recipient can independently share a message to all of their followers, is the architectural feature that enables political movements to organize and misinformation campaigns to find traction and shape politics.

Architecture’s Impact on Markets

Twitter has recently gone private, so it is no longer beholden to public stakeholder interests. However, the $44 billion purchase of Twitter raises pressure to find compelling ways to monetize its users for positive revenue streams. Twitter has begun experimenting with monetization strategies including its paid Twitter Blue user and account subscriptions. The architecture of Twitter can be found in the network design of alternatives like Mastodon and Bluesky, which are now more prominently competing within Twitter’s niche.

The eight security cards corresponding to human impact are:

• emotional well-being
• financial well-being
• physical well-being
• societal well-being
• personal data
• relationships
• the biosphere
• unusual impacts

For the scope of this values redesign, I’ll center my attention on emotional well-being, societal well-being, and relationships. Roughly speaking, my alternate values are built for connecting people, promoting community, and healthier interactions, and these three cards relate to our interactions and well-being.

One potential design change adhering to these values could be to limit visibility of metrics. LinkedIn’s designers made a conscious decision to obscure exact ‘following’ metrics by rounding the connections displayed on the profile and capping the number at 500. Similar logic could work for Twitter – by reducing clarity of metrics, we eliminate the extrinsic incentives for making numbers go up. Examining this through the lens of relationships and societal well-being, the elimination of metrics encourages users to use the platform without the expectation of being rewarded through numbers. This defers the positive feedback expectations towards comments.

Another design change, which is potentially controversial, may be to limit one’s reach for ordinary (non-media) users. There are a few ways this could be done. One approach might limit reach to within two degrees of a user’s followers. Another may cap the number of views a tweet can receive from unique users (i.e., only a maximum of 1000 people may see your post). Regardless, limiting reach serves to change how users think about their audience. By containing reach, we also limit the spread of misinformation. Users will need to reframe their perspective – rather than encouraging them to broadcast their thoughts to the world, what’s important is to express your thoughts with the people you’re around.

Redesigning is difficult! I think my approaches to limit metrics and reach to reprioritize user incentives are somewhat unorthodox, and I can equally imagine a myriad of potential downsides and loopholes that people may exploit. People could take screenshots of the original tweet and repost it to expand its reach. Web scraping and the Twitter API could be used to create third-party tools to uncover metrics like followings. Furthermore, while I’ve been focusing on the average user, the dependence of many influential figures and news publications on Twitter’s platform deserves an entirely separate consideration of needs and constraints. On the other hand, it might be a good thing that fewer people rely on Twitter to get their news. Using Twitter instead to facilitate connections within the first and second degree may promote healthier relationships and user well-being. What’s evident is that there are a lot of competing stakeholders to consider, and a redesign of Twitter from media to network leaves open questions about the niche of different groups on the platform.

1. What Things Regulate, Lawrence Lessig (from Code v2.0) ↩
2. http://securitycards.cs.washington.edu/ ↩