return ✕︎

Media

By E. Glen Weyl, Audrey Tang and ⿻ Community

Media

Immersive and telepathic media experiences promise to transform connection across difference, making the experiences of the marginalized as palpable to us as those of our neighbors. Collaborative journalism promises to increase by an order of magnitude the number of citizens who can meaningfully contribute to shaping our shared narration of history as it happens. Cryptographic securing of sources can increase freedom of the press, the equivalent of moving every country up a category (viz. from Satisfactory to Good in the Reporters without Borders World Press Freedom Index) by lessening the trade-offs between source confidentiality and state secrecy. Creating a more ⿻ structure of attention allocation and business models to support it could at least undo the rise in affective polarization in many jurisdictions and possibly reduce them to the levels seen today in the least polarized jurisdictions like Taiwan and the Netherlands.


The direct experiences most of us have in our everyday lives exposes us to only a tiny sliver of global affairs. Almost everything we believe we know beyond this is mediated through relationships, schooling and, most of the time, "media", especially journalism (radio, television, newspapers) and social media, as well as directed small or large group communications such as email and group chats. An important promise of digital technology has been to transform media, a possibility we take up here with keen awareness of the dangers and harms to media that are widely attributed to digital technology and social media. We explore how ⿻ could help correct many of these harms and help achieve something of the potential that internet pioneers like J. C. R. Licklider and Robert Taylor saw in digital media.[1]

In particular, we highlight how the coming tide of ⿻ may help increase empathy across social distance even more dramatically than photography and television did; how it could increase by an order of magnitude or more the number of people who can meaningfully and helpfully participate in the journalistic process; how it could help restore level of trust in media, as well as norms of respect for confidentiality, much of the way towards what they were at their mid-twentieth-century peak; how they could they could undo most of the rises in levels of "affective polarization" (viz. dislike across lines of political division) not just within national polities but across a range of other social organizations; and how it could help restore sustainable and aligned funding for media. In short, we show how ⿻ can help address and reverse many of the crises media face today.

Walking in others' shoes

As noted above, a central role of journalism is to allow people to experience the events and sensations of parts of the world they may never visit. Every generation of technology has made this more vivid and thus created a "smaller world". Abolitionists like Frederick Douglass harnessed photography to bring the experience of slaves to Northerner whites.[2] Radio helped make the Great War a truly World War by allowing the sounds of conflict to echo around the world. Television allowed millions to share Neil Armstrong's landing on the moon.

Immerisve shared reality promise to create even deeper empathetic connections. Journalists may soon be able to bridge social divides with vivid empathy as never before. While they have reached a limited audience so far given the image quality and nausea-related challenges of existing virtual reality (VR) headsets, journalists and artists have already begun to pioneer a variety of empathetic VR experiences. Examples include Milica Zec and Winslow Porter's work to help people experience life as non-human life like a tree, Decontee Davis's portrait of one of the world's most horrific diseases from the eyes of an Ebola survivor and Yasmin Eyalat's animated immersion within the world of cyber-security.[3]

Yet these are only the first successful forays into an emerging medium. As shared reality technologies branch out into other senses (smell, touch and taste), far more complete multisensory connections will be possible with even more surprising and enlightening results. Brain interfaces will be transformative in a way that is hard to even describe. The future of journalism empowering us to know things that are profoundly different is therefore bright.

Citizen co-journalism

One of the most important trends in the production of journalism in the internet era has been the rise of so-called "citizen journalism" and the allied "open source intelligence" movement, both of which aim to empower a much broader diversity of people than those traditionally employed as formal journalists or intelligence analysts to document important events in the world around them. Such journalism has been central to documenting many of the most important events in recent years, from terrorist attacks to wars and police abuse. Yet it also faces significant criticism and social concern over bias, rigor of verification of facts and legibility and digestibility.

It is easy to see how many recent technological developments could dramatically exacerbate these problems. Generative foundation models (GFMs) will make the production of realistic fakes far easier and will spread distrust of any material without rigorous, multi-source validation. The echo chambers of anti-social media will allow fakes to spread even absent such vetting, proliferating misleading content and the conditions under which people will believe it.

Yet there are equally clear precedents for how technology could offset these challenges. Wikipedia has shown the speed and scale at which distributed participation can produce roughly and broadly consensual accounts of many events, though not quite yet at the speed required of journalism. Many of the tools we have described above and detail below can help address challenges of rigorous verification at distance and scale and rapid achievement of rough and socially contextual consensus that is a more appropriate frame for thinking about "objectivity".

Perhaps one of the most interesting possibilities, though, is the way in which GFMs may allow for a new form of coherent, digestible, broadly traveling and yet authentic community voice. There is a long standing tension in journalism between allowing a community to "speak for itself" (often through quotes or extended descriptions of community practices) and crafting a compelling narrative digestible to the target audience, and an even greater one that arises when articles are translated for other audiences. GFMs will increasingly allow communities to finesse these trade-offs, as they can learn from and synthesize the speech patterns of community members, incorporate verified facts and at the same time smoothly translate to a range of languages and subcultural standards and styles. This will empower groups of citizens who are not trained as journalists to convey the important stories they have to tell with precision and clarity to diverse publics.

Cryptographically secure sources

One of the most frequently dramatized tensions in journalism surrounds the role of source confidentiality. Confidentiality to the subject of the report is often broken by a confidential source to create the credibility of reports. Journalists have to verify the authenticity of their sources and the information they provide, while ensuring their secrecy from (among others) the organizations they inform on and the credibility of their report to the public. In many cases, confidential informants are sharing information that the norms of their organization prohibits them from sharing. This creates strong tensions between many of the values we have highlighted above: of protecting the associations, ensuring the integrity of the public sphere, etc. How might the tools of ⿻ help navigate these challenging waters?

Many parts of the of the above process are naturally facilitated by the tools we highlight in the "Identity and Personhood" and "Association and ⿻ Publics" chapters. Most of the tools for protecting ⿻ publics could be applied by organizations to reduce the credibility of documents shared outside their intended social context. At the same time, zero-knowledge proofs (ZKPs) based on public credentials could allow sources to remain confidential even to journalists while proving (elements of) their position to journalists' audiences. Yet, absent some reconciliation, such strategies could quickly become an "arms race", escalating cryptography without arriving at a better social outcome.

A potential resolution of this impasse arises from the subtle distinctions these protocols make regarding verification. If someone publicly holds a position in an organization, they will typically be able to prove this to others using a ZKP without revealing other elements of their identity. They may be then able to harness the associated reputation, but no more, to make claims about things occurring in the organization. But for more sensitive information and expansive claims, especially if the person only holds a relatively low position in the organization, additional verification will usually be required to make this credible. One way is by revealing more (public) information about themselves, but this will narrow the pool of people they could be and thus expose them. Another is to provide direct verification ("receipts") of the claims. However, if these receipts are protected by technologies like designated verifier signatures, this is only possible by exposing their "private key" to another person (e.g. the journalist or legal authority), which puts them at risk of exploitation or exposure by that other person unless she is herself highly trustworthy.

Of course, the precise details vary greatly depending on which precise tools are used by each participant in this dance. But overall, it illustrates how ⿻ cryptography can simultaneously allow for a quite intricate mix of trustworthy and private disclosures, protection of community norms of confidentiality and the ability to override these norms at personal cost in a broader social interest when critical.

Stories that bring us together

While many Americans look back with nostalgia on the history of the press, the era of "press responsibility" against which they judge the harms of anti-social media, dates only to the 1940s. This was when the "Hutchins Commission on Freedom of the Press" developed a code of social responsibility under which the press would act as the "common carriers of public discussion", creating a baseline of shared understanding on which public debate could proceed.[4] That commission argued that a central role of a free press in a democratic society is to clarify to all citizens both the points of consensus (viz. the "Walter Cronkite effect" of commonly watched, consensual news) and fact and those of divergence (viz. the "fairness doctrine" and practice of balancing diverging perspectives) to allow self-government to thrive. While many appreciate what this era achieved at the national level for one country, the essence of ⿻ is that we live (especially today) in a much richer and more diverse world, with many loci of democracy across, between, within and beyond nations. Whatever the many failings of social media, one thing it has achieved is to allow this diversity to shape the media ecosystem. How might it do this while still being pro-social media in the sense of the Hutchins report?

Our Augmented Deliberation chapter above suggests a natural strategy. Social media algorithms could create "communities" based both on patterns of behavior internal to the platform (e.g. views, likes, responses, propagation, choices to join) and on external data such as social science or group explicit self-identification (more on this below). For each such community, the algorithms could highlight "common content" (commonly agreed facts and values) of the group that span the divides internally, as well as important points of division within the community. Content could then be highlighted to members of the communities within this social context, making clear which content is rough consensus in communities that citizen is a member and which content is divisive, as well as offering opportunities for the citizen to explore content that is consensus on the other side of each divide from the one she is on within that community.

Such a design would continue to offer individuals and communities the agency social media affords them to respectively shape their own intersectional identities and self-govern. Yet at the same time it would avoid the rampant "false consensus" effect where netizens come to believe that extreme or idiosyncratic views are widely shared, fueling demonization of those who do not share them and a feeling of resentment when associated political outcomes are not achieved or "pluralistic ignorance" where netizens are unable to act collectively on "silent majority" views.[5] Furthermore, and perhaps most importantly, it would reshape the incentives of journalists and other creators away from divisive content and towards stories that bring us together. It is relevant beyond "hard journalism" per se as many other cultural forms (e.g. music) benefit from audiences who want to share cultural objects and fandom with others.

⿻ public media

The recommendations of the Hutchins Commission were largely adopted by leading media outlets as part of the then-prominent campaign for "social responsibility", which has recently made a comeback in the form of commitments to "environmental, social and governance" (ESG) goals among many companies. Yet a firmer foundation for encouraging such responsibility would be to align the funding sources of media more closely with the pro-social design goals above.

Neither individual subscriptions nor advertising offer a particularly promising path here, as both aim to appeal to consumers rather than citizens of diverse communities, and thus encourage serving consumers only the "dessert" they are tempted by rather than balancing this with the "vegetables" that bring them together with their communities. If we want social media to bring us together, we should aspire for it to be funded by organizations with a dedicated interest in achieving that goal: collective organizations including churches, civic associations, governments at many levels, charities, universities, corporations etc.

Replacing advertising with funding from a diversity of communities does not require much of a stretch of imagination from existing business models in adjacent industries. One of the largest and most profitable business models, pursued by corporations like Microsoft and Slack, is selling productivity software, which often includes social media-like components, to companies to boost productivity. These companies have no interest in "engaged" or polarized employees; the goal of the tools is to bring employees together to accomplish shared goals and adjust to change. A new, pro-social media model could thus naturally be incubated in such settings and then sold, in broader social contexts, to other organizations interested in solidarity and dynamism.

Furthermore, there are good reasons to believe such organizations could afford to displace advertising revenue. Most democratic governments (e.g. Germany, Finland, United States) spend more than a billion dollars a year supporting public media and far more than that subsidizing other culture.[6] Even religious media received more than $100 million in the United States alone in 2022.[7] This compares to roughly $5 billion in advertising revenue earned by Twitter (now X) in 2022, at its peak.[8] It thus seems quite plausible that, together, a range of community representative organizations could replace advertising as a revenue stream, if community leaders focused on this space and social media channeled its attention to this new business model.

This might play out in a variety of ways, but a simple one would be for participants to opt into a set of communities they identify with. Each would "sponsor" their community members' use in exchange for the prioritization for their members' attention to the community-relevant content we discussed above. Users who did not sign up for communities paying sufficiently might have to accept some amount of advertising or pay a subscription fee, and the service could identify from its own patterns communities and approach their leaders to ask for payment. In short, social media might become a more ⿻ version of public media.

Overall, the examples above show how ⿻ can empower a new pro-social, ⿻ media environment: one where we can connect deeply with others from very different backgrounds as us, where people come together to tell their stories in authoritative and verifiable ways without compromising community or individual privacy and where we come to understand what unites and divides us in the interests of the dynamism and solidarity of all our communities.


  1. Licklider and Taylor, op. cit. ↩︎

  2. John Stauffer, Zoe Trodd, and Celeste-Marie Bernier, Picturing Frederick Douglass: An Illustrated Biography of the Nineteenth Century's Most Photographed American (New York: Liveright, 2015). ↩︎

  3. Milica Zec and Winslow Porter, Tree (2017). Decontee Davis, Surviving Ebola (2015). Yasmin Elayat, Zero Days VR (2017). ↩︎

  4. The Commission on Freedom of the Press, A Free and Responsible Press: A General Report on Mass Communications (Chicago: University of Chicago Press, 1947). ↩︎

  5. Gary Marks and Norman Miller, "Ten Years of Research on the False-Consensus Effect: An Empirical and Theoretical Review, Psyhcological Bulletin 102, no. 1: 72-90. Deborah A. Prentice and Dale T. Miller, "Pluralistic Ignorance and the Perpetuation of Social Norms by Unwitting Actors", Advances in Social Psychology 28 (1996): 161-209. An example of false consensus is that many observers believe SARS-Cov-2 escaped from a laboratory ('lab leak' hypothesis). The rationalist web site Rootclaim even assessed 'lab leak' at 89% probability (~8 to 1 in favour). Subsequently, educated laypersons were exposed to the evidence in over 18 hours of adversarial debate and found posterior probabilities on the order of ~800 to 1 against lab leak, implying a Bayes factor of ~100,000 to 1 against lab leak. Despite the strength of the evidence, the lab leak claim persists since not only does zoonosis lack emotional resonance but it also requires hard work to evaluate and offers no cathartic pay-off. Similarly, due to pluralistic ignorance, despite the fact that more than 81 million people in the United States voted for Joe Biden in 2020, a small crowd of several thousand highly motivated individuals almost succeeded in disrupting the Electoral College vote count on 6 January 2021. Jonathan E. Pekar et al., "The Molecular Epidemiology of Multiple Zoonotic Origins of SARS-CoV-2", Science 377, no. 6609 960-966. Michael Worobey et al., "The Huanan Seafood Wholesale Market in Wuhan was the Early Epicenter of the COVID-19 Pandemic", Science 377, no. 6609: 951-959. ↩︎

  6. Kleis Nielsen, Rasmus, and Geert Linnebank, “Public Support for the Media: A Six-Country Overview of Direct and Indirect Subsidies,” (Oxfordshire: Reuters Institute for the Study of Journalism: University of Oxford, 2011), https://reutersinstitute.politics.ox.ac.uk/sites/default/files/2017-11/Public%20support%20for%20Media.pdf. ↩︎

  7. “Grants for Religious Media Organizations,” Cause IQ, n.d., https://www.causeiq.com/directory/grants/grants-for-religious-media-organizations/. ↩︎

  8. “Advertising Revenue of X (Formerly Twitter) Worldwide from 2017 to 2027,” Statista, 2023, https://www.statista.com/statistics/271337/twitters-advertising-revenue-worldwide/. ↩︎