This was originally written for a class Spring 2024, in response to Zeynep Tufekci's Twitter & Teargas, up to chapter 2. It resurfaced while I was looking through the writing I produced during grad school, and it's a fun window into what I was thinking about the year before PMsky really materialized. I didn't start working on that platform until the fall, but we can see here the ideas that were percolating months prior.
Tufekci's book starts by documenting the rise of social media as it overtakes traditional media as a source of truth. This piece builds upon the ideas of 140journos, an independent citizen journalist collective, with a goal of rapidly spreading information about the physical world digitally, attempting to balance verification with censorship.
The Problem of Information and Censorship
Tufekci mentions a few times that the information onslaught of the internet is "somewhat chaotic", which is true. They document the shift in Turkey, and elsewhere from Orwellian censorship where information is blocked, to Huxleyan censorship, where information pathways are overwhelmed, diluting information to the point that viewers can't tell what's fact and what isn't.
Many outlets of today's internet eschew censorship, proclaiming "free speech", and ending up dealing in Huxleyan onslaughts of sludge. While a lack of top-down, centralized censorship is an improvement, it's far from a perfect landscape for media, communication, and information. So the question is, how do we terraform that landscape to surface verified facts, without giving too much power to any central "verifier", backsliding into Orwellian censorship once again?
Philosophy of Truth
Truth itself is a nebulous concept, with much philosophical thought dedicated to uncovering what it really is. Correspondence Theory claims that the truth of a statement is related to how well it corresponds to reality. This is nice, but who gets to be the authority on what reality is? Several other theories, (consensus, coherence, pragmatist, and constructivist views) take a more relativist stance on truth - that we construct that reality collectively. Additionally, deflationist or minimalist philosophers claim that truth is naught but that which we assign to truth.
If we try to assert that information on a platform follows Correspondence Theory, i.e. what you find on the platform is True in reality, we'll have a hard time proving that. Moderation of this kind doesn't scale; unless a verified profile was there in person, we won't have 100% certainty.
In conceiving of a platform whose goals are dissemination of true information and avoiding centralization pitfalls, we must consider the other strands of thinking on what truth really is. Claiming that any given post is 100% "real" requires some authority on reality. What if instead, we approached truth from the Coherence or Constructivist perspectives?
Sidebar on Liquid Democracy
The current political system in the US, and in many other places, is a representative democracy. The public elects representatives, who then vote on policy, theoretically voicing the opinions of their constituents. This is a vestige of a pre-internet age, where information traveled too slowly to get everyone in on the conversation. With the internet, information can travel quickly enough, but as Tufekci points out, our new bottleneck is attention.
The concept of Liquid Democracy aims to solve this through a few minor changes to our representative system. First, anyone can vote - in our current day, there's nothing stopping additional participation. Second, if you don't vote, your say gets funneled to your representative - who you can choose, and change at any time. These two changes modernize our antiquated representative system, allowing for far greater participation, while still allowing people to check out of the political system. That is, until something draws their attention in, at which point they can vote for themselves or change their representative in minutes.
We can view news and the dissemination of information in parallel to the problem of politics and policymaking. Previously, only an authority could disseminate information (traditional media, representatives). Now, anyone can post on social media (vote on policy), and curators (new representatives in a liquid democracy) share information, their reach associated with their credibility. Similar to how liquid democracy would revolutionize our policy making, social media and curators have revolutionized media.
Proposal for a New Information Platform
With those concepts in mind, let's envision a new platform for the dissemination of media. Here's our main requirements:
Minimal, or no censorship. no one person or entity can block any other from publishing information.
For a regular person, the information they see is relatively true. generally credible information bubbles to the surface, while generally incredible information doesn't make it very far.
Tufekci mentions,
> As technologies change, and as they alter the societal architectures of visibility, access, and community, they also affect the contours of the public sphere, which in turn affects social norms and political structures. (6)
Our goal is architecting a space where information flows freely. One where truth is rewarded with a wider reach, and "fake news" is quickly labelled as such, in a distributed fashion.
Inspiration might come from Twitter's "community notes" feature, or Bluesky's explorations in decentralized moderation. We can also look to Holochain's paradigm shift from data-centric applications (where the source-of-truth is a company's datacenters) to agent-centric, where each user's device keeps track of what they "know" about the state of the network.
Architecture
In order to avoid the pitfalls of having one legal entity in charge of censorship on the platform, this technology will take the form of a protocol, like email. Data could be stored by entities on different continents, but also on platforms like IPFS, or it could be held locally and replicated peer-to-peer, like technologies such as bittorrent, holochain, and others. The goal here is that posting doesn't rely on a potentially-corrupted outside source to host what you deem worthy of sharing.
Verification
Once we sidestep Orwellian censorship, Huxley's vision quickly comes into fruition. The question here is how to keep track of credibility. Current social media have a "verification" process, where the platform itself decides who is an "official" account. This platform however, will have no such central entity. Credibility will be entirely relational. Rather than "This is the official account," the statement will be "Twitter.com says this is the official account".
Tufekci details the deep networks of relationships these citizen journalists and curators have - the idea is to create a platform where these relationships form the bedrock of the information network.
Consider the example of Ahmed Omran, who was able to curate and fact-check digital posts remotely. Their expertise allows them to verify new information. What if their thought process were public, and collective?
Over time, as bonds grow, mutual credibility grows as well. By integrating photo and video metadata, we can multiply the effects of these curators. Posts that are collectively verified gain steam, while perspectives deemed questionable (again, collectively) are marginalized.
We can allow for linking of posts, so that one can view an event, and see the network of information related to it. As posts interweave themselves, the plurality of perspectives begin to surface a more and more likely view of reality. By implementing a mixture of Reddit's karma system, Twitter's community notes, and Bluesky's moderation system on a distributed network, citizen journalism could reach new heights, soaring above even today's versions of censorship.
On the surface, such a platform might look and feel similar to Twitter, and other microblogging sites with familiar interfaces, albeit with a few new modals for more detailed interactions. Underneath though, on the backend's servers, the algorithm deciding what posts you see when would be geared towards veracity and importance, rather than driving clicks, revenue, or ad-spend.
Conclusion
In summary, the goal of this intervention is to re-examine the platforms that make up our digital media landscape. In reassessing our goals of sharing and verifying information, how can we configure a platform to encourage the spread of good information, while stifling Huxleyan noise? Given the technologies we have today, is a better media platform possible? How do we measure truthiness? What metrics can we introduce and utilize? What shape of digital landscape lets users gather at consensus and truth, rather than polarization and division? How can we achieve this collectively, without a benevolent dictator at the wheel?