Fixing Social Media Part 1: The Big Fix

I'm going to talk about a topic which isn't usually what I blog about, here or anywhere. It's on my mind, though, and I think it's extremely important, even more so after reading this article in the Washington Monthly on the major issues with how Facebook operates and how that allowe bad actors to influence events such as the Brexit vote and the US election - to their (perceived, short-term) benefit and the rest of the world's detriment.

I'm doing this in two parts. The first is about what I think could be a fix to at least some of the issues mentioned in the article. That's not something I could implement myself (or rather, it's not something I could implement myself and make functional enough and, crucially, widely adopted enough for it to replace the current social media model); so I'm also planning a Part 2, on what I am trying to do, and what you can also do, on an individual level to mitigate the current problems with social media and make it ever-so-slightly less awful. I'm putting this out there, though, in case anyone who can make it happen wants to pick it up and run with it.

As a background, my day job is as a business analyst for a large corporation, and I've spent 20 years designing technology solutions for real-world problems. I'm not the best-qualified person in the world to suggest how to fix social media, but I'm also not completely without a clue. Here are my thoughts.

Opening Social Media

Most of the current problems with social media, as identified by the Washington Monthly article, occur because it's not controlled primarily by the users, but by largely unaccountable and almost unregulated corporations.

The heart of my proposal is to change this, and the means isn't (essentially) new. I propose a social media landscape built forward from the technology we had around the turn of the century: blogs and RSS.

The idea of the "blogosphere" was that people had a conversation by posting their thoughts and their responses to others' thoughts. But blog platforms never really facilitated that as well as they could have. Social media platforms make it much easier, because they're centred on the social network rather than on the content. However, the key thing about blogging that's different from social media, and better, is that you can own the content and host it on a site that you control - though you can also use free platforms that will host it for you if you prefer. This blog post, for example, is coming to you from my own site; I own the domain name and pay for the hosting, so I can pick it up and take it somewhere else if I want to.

RSS and its various successors are technologies by which you can subscribe to the entries on someone's blog (or similar) and see them turn up in a "feed reader". Google used to have a good one, but discontinued it, probably because in the age of social media fewer people were following blogs. I migrated to Feedly, but I don't keep up with the feeds much anymore either; I mostly just use it to make sure I don't miss any episodes of webcomics I follow. This is partly because it's not in a single stream, though I can put it into one if I really want to, combining multiple feeds under a topic that I define. Still, it doesn't produce a social-media-like experience.

There was an attempt, years ago, to create a platform that aggregated your various feeds from all the different sites you contributed to into one place where people could subscribe to them all. I don't even remember what it was called; the implementation wasn't compelling, and not enough people adopted it. But the idea of a unified feed that would show anyone who cares your Twitter, Tumblr, Instagram, Goodreads, etc. all in one place is a good one. Google+, Twitter and Facebook facilitate this, in fact, by enabling you to autopost from other platforms.

So here's what I imagine as a compelling alternative to the current monolithic social media platforms which are in control of the content you produce, the content you consume, and your social network. It combines those older ideas, and brings them up to date.

1. A new, open standard for "feeds" that can take content in text or multimedia forms, posted on any of an arbitrarily large number of platforms that conform to the standard; tag it in various ways that the user controls; and combine it in a way people can subscribe to.

The crucial thing about this is that the content remains on whatever site you posted it on. You absolutely can still post on Twitter or Facebook or Google+, but you can also post on your own hosted site, or Goodreads, or a completely new site that someone invents tomorrow, and (through a small snippet of code that anyone can include on their platform) have a checkbox or whatever that says "add to my feed". And anyone who subscribes to that feed can see it, because they're subscribing to the whole feed, not an algorithmically curated version of the feed controlled by an unaccountable platform. Nor do they need to be using Facebook to see your Facebook posts.

Anyone can create an app that allows people to post to their feed, and anyone can create an app that allows people to view the items in the feeds they're subscribed to. If Google+ changes their app in a way that annoys you, you don't have to stick with it to see content from Google+ users. You're in control of both your content and your user experience.

A crucial upgrade to the old feed system to make it more social-media-like would be the ability to post comments on the content straight from the feed reader app, in a similarly open and interoperable way to the posting of the content to begin with.

Of course, you will want to curate and filter the feeds you subscribe to. I'll get to that in a minute. First, how would you choose to subscribe to someone's feed? That's the next part of the solution:

2. An open social graph (which is what the connections between people in a social network get called by people who discuss these things).

It should be straightforward and trivial to provide, on whatever site you happen to be posting on, a subscription button that will allow people to link into your feed and select what parts of it they want to see - more on that below. When they do so, they should be presented, in whatever app they're using, with the option of categorising their connection to you: family, social, education (teacher, student, fellow student), employment (colleague, former colleague, employer, union), geographical (neighbour, old roommate, local organization), interest (religious group, fandom), and so forth. This is in their app and under their control; it isn't owned by a corporation or exposed publicly.

"Free" apps might have as part of their terms and conditions that they can make use of what they know about your connections for advertising - ideally revealing that they are doing so as they do so. Also, there are obvious advantages to my app and your app exchanging this information; if you and I went to school together, I'm likely to want to know about the people you're connected to who also went to school with us, and connect with them as well. But that doesn't require that the information be publicly accessible, or owned by a corporation. It can just travel between our apps in a private message.

Those two elements by themselves would be adequate to get the system going. To make it compelling, though, might require a little more.

3. An open topic system.

One thing that Google+ has over the other social networks is that you can self-categorize your content into "collections", and people can decide whether or not to see it based on that. If you post about robots and your dog, and I'm only interested in the robots and your sister is only interested in the dog, we can arrange it so that we only see the posts we're interested in.

There are limitations to the "collections" system, though. A post can only be in one collection; it's alternative to, not integrated with, "circles" (the way G+ provides to group people you're connected with); and the collection titles are chosen by the user and not related to any standardised list of topics. So I have a collection called "Every Day is Caturday", my friend Daniel has one called "Look at My Cats", and my friend Jen has one called "Because the Internet Doesn't Have Enough Cat Photos". All three are on the same topic: pictures of our respective cats.

A picture of my cat.

I'm envisaging a topic hierarchy, with some level of automation based on the keywords you use in a post, what a picture is of, and where you're posting, so a review of a fantasy novel on Goodreads would automatically get topic-tagged as Arts and Entertainment > Fiction > Fantasy Fiction and Arts and Entertainment > Fiction > Novels. Yes, multiple tags should be possible. Also, it would be good to be able to direct posts on a particular topic primarily or exclusively to a group of your connections, so that my political posts don't enrage my wife's family, for example. (How to implement making posts fully private, rather than just hiding them by default, is a problem I haven't figured out yet.)

Though not essential to an open social media system, I think a more robust topic system would help to get people who are interested in the same things together; to allow people to curate which posts they see - manually or with automated help from whatever app they use; and even, perhaps, to counter the "filter bubble". For example, your app could tell you not only what people you're connected to are talking about in a topic you're interested in, but also what people you're not connected to are talking about.

Further on curation, I can imagine an approach which bubbles the most engaging posts to the top, but also shows you that there are other posts from this user, or other posts on this topic, that are below some "interest threshold" that you've set. If you're feeling like looking at lots of cat photos today, and you're not so much in a dog mood, your app could cater to that.

Another potential fitering approach would be by sentiment. Sentiment analysis on social media is already well developed, based both on the words used in the post and on people's reactions to them. Corporations use it to get alerted when people are unhappy with their products or services. What if you could set your app to show you mainly hopeful posts, and to suppress anxious or negative posts? True, you could end up living in a delusional world of happy rainbows and ignoring real, urgent problems, but there are ways around that as well (if a negative post gets a lot of engagement, show it even to people who don't want to see many negative posts).

Right now, with the apps through which we access social media being controlled by the same corporations that control the content, what I've just outlined isn't an option. We're stuck with what their algorithms show us. But if the system was open, and anyone could write an app for consuming content that anyone could write an app for creating... Well, I think that would be a better world.

While we're waiting for that better world, though, there are things we can do individually to make social media less broken. More of that in my next post.

Spread the word

Leave a Reply

Subscribe without commenting