Decentralizing the Social Media/Network Landscape

Ben Goertzel
SingularityNET
Published in
6 min readDec 23, 2022

--

For better and worse, social media networks and associated technologies are now playing a key role in the thought and feeling process of our species’ “Global Brain.”

The positives of modern social media tech and culture are dramatic and easy to take for granted now that we’re thick in the middle of them -

  • We’re able to get rapid feedback on new ideas or thoughts from diverse individuals across the globe — then incorporate this feedback into our thinking and get a new iteration of feedback — enabling emergent collective thought processes on time-scales of minutes, hours, days, or months depending on the particulars.
  • We can find and form communities of people knowledgeable and passionate about obscure technical, conceptual, or artistic topics — whereas in the pre-Internet era, one could literally spend one’s whole life never encountering anyone else sharing one’s peculiar obsessions (to draw some random examples from my own work and life — applications of category theory to cognitive science, neoclassical metal violin, baking with nut flours, programming in dependently typed languages, unschooling, Taoist dream yoga… before the modern Internet, how hard would it have been to find other people deep into such things?)

The negatives of modern social media tech and culture are also fairly extreme and tend to get more airtime lately -

  • The culture of brief snippets and rapid feedback often seems to militate against complex original thinking that takes time, reflection, and study to absorb.
  • The ability to “find one’s people” also leads to new forms of tribalism, in which one has such an easy time connecting with those who already agree with one’s opinions that one rarely ventures outside this bubble.
  • Commercial and political organizations, and occasionally giftedly demagogic individuals, have understood these aspects of the social media landscape extremely well, and have become expert at leveraging them to manipulate and “program” peoples’ minds in accordance with their own interests.

The current result of the superposition of these negatives and positives is a sort of idiot-savantic global brain, with rampant misinformation and reptile-brain stupidity as the default situation, interrupted or layered here and there by pockets of incredible synergetic collective-insight brilliance.

It has been clear to me since well before the advent and broad adoption of modern social networks that the only way to create a social-network sphere in which the negatives are mostly decorations on the positives, rather than vice versa, would be to create a fundamentally decentralized infrastructure for every aspect of social media, from the basic transmission of messages to the management of identity and history to the storage of data and execution of AI algorithms, and etc, etc.

This deep need for radical decentralization in the social media domain is not primarily for technical reasons but rather political, sociological, regulatory, and economic ones. The Big Tech companies that dominate the Internet landscape today are in the businesses primarily of content moderation and ad placement, and secondarily user interface design and infrastructure maintenance. The pitfalls of bundling ad placement so tightly with content moderation are obvious, as are the general difficulties of saddling any specific legal entity with the task of content moderation across all the extremes of human belief and practice relevant to all domains of human pursuit.

Any commercial social media company has no rational choice but to make content moderation decisions based on a mix of fear of lawsuits or regulatory action, fear of alienating advertisers, and fear of losing users to other networks. But having a public social sphere sculpted centrally based on all these fears is a recipe for conformism and stupidity — these factors inexorably militate toward social networks that give a core of ad-clicking users content and experience lying in the intersection of what they want and what governments and advertisers will most easily tolerate.

This general line of thinking regarding the intrinsic weaknesses of centralized social media has become more prevalent in the years since my previous SingularityNET blog posts and media on this topic (Link 1, Link 2, Link 3, Link 4, Link 5). Elon Musk has now bought Twitter, partly with the goal of loosening up the platform’s content moderation a bit — but without giving much evidence of a deep understanding of the issues and dynamics that have made modern commercial social networks the perversities they are. Twitter founder and ex-CEO Jack Dorsey has launched the AT Protocol, a reasonably well-thought-out decentralized protocol that makes a start toward a tech stack realizing a decentralized-alternative social media sphere. But AT Protocol, nice as it is, only touches certain portions of the problem at hand.

Spurred in large part by discussions with Nick Nayfack, a fellow participant in the Cardano ecosystem, I spent a little time earlier this year writing down my thoughts on the various issues confronting those who would revolutionize the social media sphere with novel decentralized infrastructure. Today, I would like to present the result of this contemplation, collected into a whitepaper titled “Decentralizing the Social Media/Network Landscape”. This is a collaborative work, created with thought-partners, feedback, and group editing from some of my SingularityNET and Hypercycle colleagues. If these topics interest you — and they damn well should — please read, enjoy, propagate and comment!

At this point, the core elements needed for a new decentralized social media infrastructure are not so mysterious. Individual network participants need to own their own data, their own profiles, their own graphs, and histories. Messaging needs to be secure and peer-to-peer. Media channels should take the form of groupings of individuals and messages, living on top of an underlying decentralized messaging/identity/data layer, and it’s at this “media channel” layer that content moderation should exist. If a certain channel rejects a given user or message, that’s fine — it still exists in the decentralized social web, but members of that channel don’t need to see it. AI should be transparent and configurable, serving the needs of network members and providing autonomous guidance only along explicitly communicated lines. Advertising should be transparent as well, with independent AIs advising participants who have opted into advertising (e.g. in exchange for “free” services) on the biases and interests of the AI bots that are choosing their ads.

But of course, each of these conceptually fairly simple points conceals its own subtly-shaped hidden rocks and loads of cutting-edge engineering work.

The SingularityNET platform, with its capability for hosting and connecting AI agents and models in a decentralized way, can form part of the solution here. As can the NuNet protocol and its decentralization of hardware infrastructure. As can Dorsey’s AT protocol. As can the Cardano blockchain with its robust scalability, and the HyperCycle sidechain with its ability to coordinate secure decentralized transactions and compute operations in a scalable and fundamentally decentralized way. Connecting all these pieces (and others) poses multiple challenges but none of them are extraordinarily large in themselves. The rate at which this will happen is going to depend less on the technical side, and more on the allocation of resources to the matter.

There are very viable business models associated with decentralized social media. Some users will pay subscriptions for ad-free social media, and others will consciously accept ads in exchange for free access, just as they do in centralized social networks but with much greater agency and transparency. The funds from these revenue engines can flow through to those providing diverse services underlying end-user functionalities. However, the logic of decentralized media networks naturally lends itself to a business ecosystem in which a relatively large number of actors profit reasonably from modest roles, rather than having a small number of companies own nearly everything and extract nearly all the profit from the ecosystem. This is not the sort of business situation that VCs tend to like, posing fundraising challenges for the decentralized social media sphere — but these are challenges that clearly are overcomable, using innovative capitalization mechanisms such as tokenomics and equity crowdfunding.

Making decentralized social networking function robustly in all needed regards and getting it widely adopted won’t be easy, but it’s a very critical thing given the high value of having an even slightly savvier Global Brain moving forward through the last few decades before the Technological Singularity. Social media can seem frivolous and silly, and a lot of the current content out there definitely is, but they are also among the core mechanisms guiding the collective perception, cognition, and action of our global society. It’s one of those things that seems overblown at first, and then you realize it’s actually weird-but-true — It’s entirely possible that the quality of our social media ecosystem could make a critical difference in whether the Singularity we get is mainly positive or negative, or even in whether as a species we get to a Singularity rather than annihilating ourselves or meeting some other unpleasant fate.

Stay Up to Date With the Latest News, Follow Us on:

--

--