Social media platforms should be required say what they are: distributors that show people what they asked to see, or publishers that decide what people should see. They have spent years avoiding accountability by being neither fully responsible for their content nor fully neutral about it, and using it to amplify opinions to their own benefit. What if you could choose a platform that respected your decisions about what to see, or actively choose one that gave you the narrative content, with recommendations and discovery as features?
We're being manipulated by companies who do not have our best interests at heart. However, the service they provide could be something fantastic. I argue that, in its purest, most honest form, social media is fantastic. We're not going to stop the transition to existing perpetually online, and so too will go our friendships, connections, and social engagement. As such, we need a framework to start talking about what is being done to the content we see.
In the USA, Section 230 shields social media companies from legal liability for content posted by their users. I do not believe it is fair to manipulate what people see while hiding behind "that is not our post." If the platform decides who sees it, when they see it, and how strongly it is pushed, then the platform has turned that speech into product. It stopped being the user's content once it was fed into the profit-machine.
There's no avoiding the harm that this continues to cause. Amnesty International reported the genocide pushed by the editorial (see: algorithmic) decisions of Facebook, documenting the extent of hate-speech pushed across the nation as "high-engagement content" spread across Myanmar in 2017. Consider also the influence on the outcome of the 2016 election in the USA, in which The Zuck's Meta was handed a 5bn USD fine for data collection, which was later used by Ted Cruz and Donald Trump. The distribution of COVID misinformation across that era follows a similar pattern.
I suggest that they can be either a publisher, or a distributor, but not both. A distributor shows people what they asked to see. A publisher decides what people should see.
As a publisher, they curate, editorialise, and create the content you see. They take content from your friends, their friends, influencers, celebrities, and internal AI-generated content. They spin together a creative narrative and deliver it to you whenever you want it. Of course, they spin ads through this too.
As a distributor, they are bound to showing exactly what is posted, chronologically, and in line with the things you've chosen to see. You don't see suggestions, you don't see celebrities, you don't see AI content. You see things as they are, with no narrative. There are, of course, some requirements for moderation, which we'll discuss later.
Publishers are not strictly bad. They surface otherwise hidden content, and have created seemingly emergent communities (see BookTok) of their own volition. They produce viral hits and bring new talent to light. Entire companies exist to game the algorithm of these publishers, and it's become a major economic driver. Similarly, it provides the opportunity for reverberance amongst dissenters and counter-culturists, who may otherwise struggle to be heard — Russian influencers rebutting Putin on Instagram, for example.
The publisher model is closer to what we have now. They are allowed carte blanche control over the content that people provide them, including edits, amplifying, silencing, and timing. They are, therefore, responsible for the impacts of what it shows to its audience. Did an underage user see inappropriate content? Then the publisher is at fault. Does the platform lean towards one way or another? Then this is entirely a decision made at the hands of the company.
In the second instance, the distributor is no longer held accountable for the content it displays. It is simply a pass-through for content that people have asked specifically to see. This encourages both active pursuit of interests by curiosity and natural exposure (a nostalgic wonder for many of the old-folk of the internet). Controls and restrictions can be implemented on the user end, either by parents, guardians, or network staff.
Under this model, distributors are still expected to engage in moderation — violent and illicit content is still removed, for example. Note that its effect would be far less inflammatory, as it reaches only a small subset of people. This is not a 4chan-anonymous place for all posts — you're still posting as you, mostly to people you know or share community with.
There are no default recommendations, follows, or friends. If there is a general push towards something, then the trend will be visible by the natural ebb and flow of what people choose to share. No post gets held above another outside of its chronological origins. The default state of a user must be nothing and nowhere, allowing them to curate a space that works for them, both initially and over time.
Moderation removes content according to public rules. Curation decides who should see it. The distinction is drawn on whether the platform is showing content as-is, or whether it's curated. The expectations around removal for distributors should be defined transparently: CSAM, violent threats, abuse, spam, impersonation, illegal material, and other categories society decides platforms should not carry. On this basis, the rules about what we choose to allow are surfaced, and rebalanced as needed.
Recommendation is a service, and this isn't serving us any more. Curation happens on two ends — when it happens on the provider side, it's manipulative and pro-advertiser. When it happens on the user side, it's a return of power to those who provide the content upon which the platform thrives.
Something that both publishers and distributors alike may want is a "front-page" of sorts. In both cases, it shows content that was not picked specifically by you, and it is not filled with things you follow. It may contain "best-of" style content, filterable and sortable. For a publisher, they can choose whatever they want to show at the "top" of this page (they're still a publisher, after all). This is also allowed under a distributor model, as the filters and selection are user curated. It would follow a standardised, public algorithm and sorting rules, with a set of pre-defined rules they could opt in or out of some of. The exact rules about what forms this algorithm can be defined later — the important point is that this page should be public, reproducible for a point-in-time, and shown equally to any user.
Scale Matters
Don't treat every platform like Instagram or TikTok. A small hobby forum is not the same as Meta's behemoth, nor is it the same as a popular but distributed forum or a group chat. The aim is to bring large platforms that choose what people see into accountability, while allowing smaller content providers to exist. It is the reach and control that is concerning, and the amplification of ideas from a single source. The measurement of scale is based on reach, userbase, and views (though I'm choosing not to pick hard numbers here).
In truth, the current swath of feed-driven companies would collapse if forced into the distributor model. I do not expect them to change. In fact, this may be an illusory choice, where no business would ever willingly choose the distributor model. This is fine — what we need is upfront expectations about what the product is. Distributors and publishers alike should be legally bound to what kind of online model they choose, and enforced following such a designation. I hope this makes it clear about exactly what they are, and provides a formal framework for how we can talk about these types of virtual places.