Last week I was happy to speak to the co-founder of the blog voyager-log, Peter Banks, and contributor Joey Yoniles (try to guess how I know him). Our meeting being short, we did not have an opportunity to exchange anything more than pleasantries and I regret that I did not get to talk to either of them more about their writings. So coming back from traveling out to meet them, I took some time to reread Joey’s first article Quarantined Idea and see how its ideas have played out in the year since its publication.

TLDR

If it isn’t obvious, you should read Quarantined Idea and all of the articles on voyagers but I will summarize its contents here. In short, the corporatization and centralization of the internet led to little niches of hyper-specific and hyper-toxic ideas. Communities deemed too toxic by the public at large are deplatformed from major sites and form echo-chambers away from prying eyes, or more likely, as golden handcuffs for struggling platform owners. These niches allow the darker aspects of humanity fester, leading to exploitation, depravity, political extremism, and abuse.

I broadly agree with the article and its conclusion: A collective duty falls on all of us to push back on ideas. Although its not any individual’s responsibility to be in the replies or enter these spaces, ratio-ing somebody spewing hateful nonsense may actually be making the world a better place.

How did we get here?

The truly wild west days of the internet were behind me. When I first ’logged on’, I wasn’t connecting to a bulletin board, Discord had long replaced IRC, and I never considered a whistle as part of a hacker’s toolkit. A lot of the primordial internet has been lost or are mirrored in tiny internet archives, often themselves a relic by modern engineering standards.

I was however a netizen just as the old web (Web 1.0) was dying off and the dynamic, and open Web 2.0 standard was taking off. Jquery was all the rage, Angular was the new hotness, and our friend React was still a library (They still say they are but thats delusional). As these new (dare I say reactive) ways of producing and managing content on central platforms such as Facebook took hold, more dispersed forms of communication: forum boards, chat rooms, etc began to disappear.

However it was not just the flashy css and ease of use that drove people to these new platforms. Social media sites provided discoverability, a promise to deliver your content to end-users, and a promise to do the same for you. The value proposition was simple: rather than setting up bookmarks, email chains, or an RSS feed, the platform would take the role of content distriubtor. Eventually you didn’t even have to subscribe to anything for the algorithms to show personalized content.

Within the walled garden of Facebook and Twitter, creator and consumer alike were offered not only an easier experience but also a more impactful one. Why publish to your personal WordPress site when Facebook has 4x the viewership and discoverability? Why sit down to publish a witty article to your email-newsletter when the tweet you whipped up on the toilet could go viral? Likewise, the promise to the platforms consumers was just as revolutionary. A netizen circa 2009 could not ‘Doomscroll’ their RSS feed or go through every niche internet forum. They had to put in the ground work to find those places first and more than likely most of the content was junk!

The only price that the consumer and content producer (who at this time would have laughed at the idea of calling themselves that) is that they had to comply with the rules of the walled garden. And of course, they had to be advertised to. In fact that’s the entire point of all of this but I digress. For creators the opportunity cost of not being shared, retweeted, reblogged, etc was too big of a price to pay. If you weren’t on Facebook in the early 2010’s you weren’t real, were stuck in the past, or were content to be excluded from the main way people were interacting with the internet.

All of this is to say that the centralization of the internet into 5 or so platforms with algorithms to maximize watch-time was inevitable. Their business model of a managed, curated internet presence is beneficial for advertisers, content-creators, and consumers alike. That business only works however if advertisers are on board with the content of the platform if investors are willing to invest in growth. However like mentioned before being with t was paid at a price and this price eventually was felt by platform providers as well.

A King to His Castle

I am the webmaster of this site but a more fitting term may be King. So long as what I am doing is legal in the US, I need only consider what I think about a piece of content before posting it. Most people do not have this luxury and have to sign on to a Terms of a Service (ToS) agreement that limits their activity to be in line with what a platform deems acceptable.

This, for the most part, is fine. Barring a crazy or uncle or two, most people in their daily lives do not produce anything close to a ban-worthy post. The main issue however still stands. Most people’s home on the internet is a little plot in someone else’s kingdom. Their presence online is by the grace of their Lord and the platform can at any time, for any reason remove a serf user from the platform. If a user is causing trouble in a kingdom, either by posting something vile enough to anger other users or more importantly scare the ad barons, then they should be removed.

Platforms of course don’t enforce these ToS agreements because their owners are prudes; they have advertisers and company stakeholders to keep happy. A hateful and vile platform not only drives away advertisers but also users who more than likely are not seeking out vile content. In many ways the ban-hammer and the ability to remove posts by a central authority is a good thing. Hate speech can be curtailed, examples of abuse like revenge-p*rn and harassment can be removed, and even somebody making an honest mistake and posting something inappropriate or embarrassing can be removed by a user themselves. A more inclusive and, I think, better internet exists today because we lived inside of the kingdom’s of Twitter, Reddit, and anywhere else the chronically online congregated.

However the impact of a company’s hand in influencing their ToS cannot be ignored. The reason that ideas quarentine themselves is because the ToS and the Kings charged with enforcing them exclude vile ideas from lingering. The drama around Elon Musk’s purchase of Twitter X only within this context. Culture and politics collided as the defining question for social media platforms goign into the new decade was answered in late November 2022 when Donald Trump was unbanned from X. But it seems that Donald was once burned and twice shy, not venturing out from his own social platform, TruthSocial, until 2023. Millions of dollars were spent on a regime change buyout of Twitter and replace who sits on the throne, but Donald Trump wanted to stay with his court on Truth Social. What I believed that most serfs users took away from the ordeal was that they did not like their new king or that it would be better if there was no king at all.

Shitposts are the common inheritance of mankind

Joey left a little golden nugget of wisdom within his preamble. The web as a global platform has the ability to create a global citizenry, a collective repository of culture, thought, and discourse. Although platforms often have cultures unto themselves, the web has failed in promoting a truly unified community. The ‘World-Wide-Web’ has fallen away leaving us just with the internet.

This is especially true as the main forms of interacting with the interenet has become centralized to a few platforms. Not only is there a chilling effect on discourse by forcing users to comply with ToS agreements but these platoforms are incentivized to clean up unsovory or vile parts of culture and prevent its dataset of ‘good’ discourse from being scraped.

A few projects carry on the ideal that Joey spoke of. Notably Wikipedia has remained community-led and maintained for decades and can be downloaded for free to preserve forever. To a lesser extent this exists for some user-generated content. Although centrally owned, Github is obviously the king of code storage and is rightfully preserved in the arctic. These two sites alone, comprising (surprisingly only) terabytes of data, serve as a facsimile of the library of Alexandra and, preserved in ice for future generations, are the common inheritance of mankind. What is missing however from this public record is a reflection of the culture of our age.

Social platforms make their money by serving ads in realtime and very recently the historical data of Facebook/X/Reddit/etc became ridiculously valuable due to the advancements in big data and LLMs. That is to say that to provide open api access or uphold the ideals of Web 2.0 would be essentially giving away millions, if not billions, of dollars. However it is exactly these datasets, produced (☝️ 🤓 and technically owned) by people at large that would be a perfect snapshot of culture and discourse at any given time. I wouldn’t hold my breath waiting for Elon to archive twitter threads in the arctic out of the goodness of his heart, especially with both how valuable they are to keep private and how voluminous they are.

Although cringe has been written into stone, the actual discourse of our time, being privately owned, is open to manipulation. Platform owners can delete and remove content as they see fit, users can (and should be able to) remove their own comments, and only God knows what happens in private channels that will never see the light of day. Not only can this data be removed, discourse and the data surrounding it, are owned* by the platforms themselves. (☝️ 🤓 ERM actually you still have copyright protection but have given the platform a non-revocable, royalty-free, and worldwide license to copy, distribute, adapt, etc your content. So they can do practically anything they want with it and often control the only way to access said content but they technically don’t own it)

In a more perfect world, I believe that a public, auditable record of culture would be kept by some public entity and preserved for future generations. Yes, your shitpost about Breaking Bad from 2019 is archived. Yes, the 40 messages that your drunk uncle left on a post made by the local parks department is archived. It should be preserved as real, pre-LLM internet discourse untouched by the personal tastes of the platforms owners. These posts are both the library of thought of the modern age and the low-background steel of digital writing. Of course, the engineering, monetary, and legal problems with pulling this off would be a stretch. Its a good thing I’m just a junior dev.

New decentralized platforms

I have been dancing my way around the open internets solution to the Kingly problem, and to a lesser extent, the problem with quarantined ideas: The Fediverse! ATProtocol and ActivityPub, popularized by Bluesky and Massive Dong Mastodon respectively promise a decentralized social platform and more user ownership. In theory, they offer a way out of the walled gardens of corporate social media. In practice, they introduce their own set of issues.

Overall, I think that these platforms are a step in the right direction but come with their own set of issues. I’m not sure if its a good or a bad thing that I am not online enough to have used either of these platforms* (* I have a threads account but that was dead within a week). Maybe as they get more mainstream I can put my money where my mouth is.

ActivityPub, the more established protocol, enables users to migrate between different instances, reducing the power of any single administrator while maintaining a federated network. It allows a variety of social media experiences under one umbrella, meaning you can theoretically follow accounts across different platforms seamlessly. And, if need be, even migrate accounts from one host to another within the federated network. Mastodon, the most well-known ActivityPub implementation, has proven this model works (at least for a certain kind of user).

Bluesky, on the other hand, overcame the biggest challenge facing alternative platforms: attracting an initial user base that isn’t composed entirely of those banned from mainstream social media. However, it has a glaring flaw: it isn’t truly federated. While Bluesky’s ATProtocol technically allows for self-hosting, running your own instance won’t connect you back to the main network. This means that, despite its marketing, Bluesky is functionally still a walled garden, just under different management.

From the perspective of a platform owner, federation means greater autonomy. An instance admin can set their own rules, ban harmful users, and defederate from instances that harbor extremism. This allows responsible admins to create safer spaces, both in the sense that they can stop hosting harmful content and block other hosts with harmful communities. However, it also means there’s no higher authority enforcing moderation across the network; deplatformed communities can simply migrate to a new instance, link up with like-minded servers, and continue operating with little consequence. In a way, instances are self-moderating and self quarantining as Joey describes but without the total isolation that comes with being banned from a mainstream social media site.

Bad actors aren’t fully erased from the network. They’re just pushed into their own corners, where extremism can fester. ActivityPub is a protocol, not a platform so its very possible to have the same toxic communities that we have in our current social media landscape, just without a king to lay down the law. Unlike a centralized platform where banning means true removal, in a federated system, banning often just means shifting the problem elsewhere.

Admins can go rogue. Harmful or even illegal content could remain with the instances of the Fediverse for decades without a central authority to enforce its deletion. Unlike centralized platforms that can enforce content moderation and deletion requests, federated networks have no guarantee that all instances will respect a user’s decision to delete or edit a post. A message reblogged across dozens of instances could live on indefinitely, with no mechanism to enforce removal.

Anybody looking at the Fediverse in a societal context should consider what harm (and legal ownership questions) these rogue instances can do. The US, at time of writing, does not have a federal right to be forgotten. Which in my opinion is a necessary component for rehabilitating and reintegrating people back into police society if they were once in a toxic community, in a harmful relationship, or otherwise need to be protected from their past.

The silver lining to that harm however is an opportunity to preserve culture for future generations. Although users still own their work, with the right legal framework a broadly federated instance may have the right to store content indefinitely. An ActivityPub instance, fully peered with multiple networks has license to store and show content from peered instances. So even while respecting the right to forgotten, a true archive of culture as it exists and evolves into the future can be legally stored internally.

Lets all touch grass again

I like grass. I like walled gardens. Whether you prefer the wild west of the Fediverse or the ecosystems built around the current social media giants, is up to you. I however believe that the Fediverse is a step in the right direction. With mainstream social media’s ease of use, and pure inertia, the Fediverse may never surpass corporate platforms in sheer popularity. It offers a more user-controlled yet curated experience and hearkens back to a time when everyone was their own king. It could also be the future that allows the general public to preserve modern discourse and discussion for future generations, not to be gatekept by CEOs training another chatbot.