I want online social networks akin to offline social networks

Ruminating more on what my ideal social network would achieve, I realize that it really comes down to one thing: Offline social networks ("meatspace", interacting with friends and acquaintances in person) are pretty amazing, and our current online social networks do a terrible job of replicating this. I'd like to capture some of what I think we're missing, without losing sight of what we also gain.

(This is a pretty rough post; I wish I had more time in which to rewrite and polish it, but some of these ideas are just burning to get out of my head.)

A slice of life

Picture yourself sitting in your living room with some friends. Inexplicably, there are no computers of any kind in the room; no TVs, phones, laptops, or always-listening female-voiced digital minions of Amazon or Google. If someone were to ask you who was listening, you could confidently list out the names of the people in the room. Not that this is weighing heavily on your mind; you're relaxed and focused on the conversation. You know these people reasonably well, and feel comfortable dipping into controversial or weighty topics based on your knowledge of how they might respond. Perhaps, based on who is present, you feel safe talking about ecological issues and reproductive rights but not about euthanasia; with another group, you might have different constraints. Maybe some of your opinions are unpopular or even heretical to the larger society, but you know that these people won't spread your comments without checking in first. If there's a misunderstanding, it can be cleared up in seconds or minutes, and you may even notice confusion, disagreement, or rising tension before the other person says anything—after all, you can see their face and posture, hear changes in their intonation. Taylor reports on his grandparent's failing health; you put your hand over his as he speaks. Blake unconsciously makes "mm-hmm" sounds as you tell them a complicated story. Alex, on the other hand, starts to get a distracted look on her face, so you wrap it up quickly. You realize she hasn't been speaking much, so you make an effort to include her more in the conversation.

Now your little group walks through your neighborhood to a café. As you step outside, your conversation becomes a little more constrained. There is less discussion of private and personal matters, since you're vaguely aware of being in a public space. Even if you can't see anyone, they might be present behind open windows or fences. At the noisy café, you feel a little more free to go into more personal topics, screened by the noise. However, you keep the topics somewhat lighter, since it's harder to receive nuance.

On the way back, one of your friends, visiting from out of country, stops to send a postcard back home. She is consciously aware that the postcard may be read by anyone in the postal service or at the destination address, and unconsciously aware that this set does not include e.g. coworkers, salespeople at the shop where she bought the postcard, or ex-partners. She chooses her tone and content accordingly.

And over here there's Facebook and Twitter

Now it's 2018 and you're elbow deep in social interactions that involve sending information to websites and then them sending that information to your friends, but also to random strangers and acquaintances. In some ways, this is cool: You can instantly talk to people you've never met before, or friends who are hundreds or thousands of miles away. But these sites that are acting as mediators for your social interactions, these social media sites, they're doing some pretty sketchy stuff, letting total randos ask questions like "give me a complete list of Taylor's friends, and all *his* friends, and all of their probable race, socioeconomic status, gender, age, political affiliation, and any other derived data". (Look, that election isn't going to swing itself!)

Even when they're not selling out machine learning models of your face, they facilitate unpleasant and unnatural situations. Someone with an axe to grind searches Twitter for references to abortion rights, finds some offhand comment you made, and gets all up in your mentions. Maybe they bring their friends, too. Or HR finds your snide remark about something negative about the industry you work in, and you get called in and chastised for what they take to be disparaging remarks about your company.

What goes wrong/what we're missing

You've probably gotten the gist of it from the preceding narrative, but just to be clear (or if you skipped over all that text), but here are some things I love about online social interactions. Except it's really hard to frame these things we take for granted in a positive sense, so I'll frame them in the negative instead. Here's what social media sites get wrong:

  • Your words are persistent, potentially saved for decades, possibly in public
  • You don't know who's listening, who your "audience" is, so you can't tailor your words and tone to match
  • Corollary of the above two: You can't know who your audience will be in 5 years
  • Backchanneling is largely absent; you can't tell how someone is taking your words unless they go to the effort of composing a response, but the effort involved might overstate their actual reaction.
  • Absent voices are hard to notice.
  • Your words are searchable; it is easy for trolls and other unpleasant parties to locate you on the basis of your identity, politics, or opinions. They can take over the conversation; if they were in your living room, you'd kick them out. Twitter has no living rooms.
  • Context-collapse is frequent/
  • Some topics are forbidden. In China, any number of topics will get your messages removed, and you might even get a visit from the police. In the US, Reddit and Twitter have recently been shutting down conversations about sex work in reaction to SESTA/FOSTA. Who would throw you out of your own living room for daring to discuss the possible decriminalization of prostitution?

What goes right

It's not all bad; people obviously use social media for a reason, and maybe not just because it's alarmingly addictive, in some cases with millions of dollars spent on enhancing that attribute. The obvious upside is the prospect of getting social time without having to leave your house, and chatting with friends on the other side of the planet. Exposure to ideas you would not have readily encountered in your own social group; a world of people who make amazing art and stories and software. Let's not lose that. And some social media software even takes steps towards remediating the above gaps:

  • On Mastodon (kind of like a federated version of Twitter), people use the "fave" button liberally; it functions not only as agreement or support, but also "I am still enjoying engaging in this conversation we are having". This is a form of backchanneling. If someone is being mentioned in a thread, but is *not* replying or faving, I will usually drop them from the mentions in my replies; I perceive them as communicating disinterest in remaining on the thread. This is a property of sites that have "reaction" buttons of various sorts that notify the recipient of who initiated the action; if the fave button only incremented a counter, it would not be useful for backchanneling.
  • Dreamwidth (the main descendant of Livejournal) makes it easy to post privately, to custom groupings of friends. Only those friends can comment on the post, but they can see each other's replies. The poster can delete or screen any of these comments. This creates a semi-private context, a custom living room. (Obviously, the Dreamwidth admins or anyone accessing a future breach of the service could read the posts or comments.)
  • Signal, an encrypted chat service, makes it impossible for the service admins to listen in; only the members of a group chat can see the messages. This is much cozier, even though the group participants could also potentially lose control of the chatlogs in the future.

We can do better

Off the cuff, some thoughts on how we can improve future platforms:

  • Posts restricted to friends, by default: Only sites seeking massive userbases to monetize need their users to post publicly and draw in the crowds. Privacy is about controlling how and when and where you choose to reveal things about yourself; Twitter and Facebook want to wrest that control from you, to increase "engagement". Social media can fight back by "going dark", largely withdrawing to living rooms and cafés.
  • An option to post publicly, when one wishes to: Sometimes you want to chat in the living room, sometimes in the café where others might join in. Sometimes you have something to say with a megaphone in the town square.
  • End-to-end encrypted by default, to provide a sense of safety and coziness; when this is in play, you have confidence in who you are speaking to.
  • Mechanisms for backchanneling: Reaction buttons, at minimum "heart" (very multi-purpose), but probably a custom emoji-like set, as seen in Slack.
  • A way to spread wonderful posts; a way to limit their spread: You might wish to set a popular post free, allow it to be shown to the world without attribution. Or just shown, with identity attached, within 3 degrees of separation. Get creative.

Responses: 2 so far

  1. Adam Snider says:

    This is great. I have a minor quibble, that is maybe not worth even posting, but I wonder if "heart" is actually multi-purpose. I prefer a star. A star can be a "like," a "fave" or even just a "save/bookmark." A heart seems to imply, at the very least, that the person likes the post in question, or even that they love it. I remember there being a minor outrage when Twitter switched from a star to a heart, for this very reason.

    I like that Mastodon and Pleroma use a star, rather than a heart. (Even better is that one of the instances I use heavily has added an explicit "save" button, that allows you to bookmark a post without alerting the author of said post. I use this sparingly, as I usually prefer to give feedback to the author in the form of the traditional "like" button.)

  2. Tim McCormack says:

    It's a good quibble! One situation where heart wins out is when someone posts something tragic, like the death of a pet. Starring such a post feels... weird. I'm having trouble thinking of examples where a heart would feel weird. I haven't used Twitter much, although I did use Imzy, which used a heart as the main feedback/backchannel mechanism. Do you have examples?

    And yeah, that conflation of "feedback" and "save this" is not great. :-)