OK, so we accept that Facebook and its employees are the Werner von Braun of Silly Con Valley, and we cannot in conscience continue to use Facebook (good riddance) or WhatsApp (who cares, we have Signal)…or Instagram.
OK, now we have a problem.
Twitter is doing some abortive cleanup work, a day late and a dollar short, but until it finds a way to eliminate all bots, it can never be cleaned up. Not least because there’s not a whole lot of granularity to it. You can set yourself private, and you can have lists to separate followers into certain categories, but I don’t think it has any mechanism for you to tweet privately to a subset of followers without resorting to DMs. And of course the priority is on everything being public. The problem is, this is not how real life works. Nobody wants to put all their business in front of everybody they know all the time. Google+ was kind of onto this, with its circles, but 1) how can you trust an ad company to build a social network and 2) they waited far too long until Facebook had critical mass and “everybody” was already there. But they built on some ideas that SixApart tried first with LiveJournal and then again with Vox – but those were blogging services, not social media as we now understand it.
I suppose it’s conceivable that you might be able to build something like this on top of Tumblr, for instance. It famously tried to split the difference around “microblogging” and ended up becoming famous mostly for porn and millennial fandom, but there’s a simple architecture there. Then there’s a new-ish app called Cocoon, built by ex-Facebook engineers – seems to be an effort at a small private space, no more than twenty participants per instance, with a steady stream for pictures and chat and even simple video calling, just a low-level shared presence. Which isn’t nothing, but it might be a little too private for what I’m thinking; it’s basically a group chat in an app. I still maintain that the group chat is the best current social media, but as opposed to Twitter or Insta, it makes it complicated to maintain multiple circles constantly. Chat is less ambient than social media is supposed to be, for that matter; I don’t get a notification every time someone I know tweets. I want to dip in and out – and that’s another thing, I want to see every post from people I follow in chronological order, not some algorithmic soup of what someone else decided should be floated to the top.
Then there’s a discovery problem. I don’t want reposting enabled, as that’s how you get viral bullshit to propagate. I also don’t want other people’s likes foisted on me; I’m tired of how Twitter dry-snitches on everything you like to your friends. (It goes without saying that the model of viral growth supported by ads is not going to be appropriate here, but I don’t know how you get around that. WhatsApp scaled on a $1 per year membership fee and that might be enough, honestly.) Basically, if I see something, it should be because I asked to see it. You should never see anything from people you don’t follow. But then, here is the problem: the weak-friendship-follow. You can’t build this on the back of something like Signal, because I don’t want to give Kara Swisher or Maciej Ceglowski my phone number but I do want to see what they say. But the model I just formulated doesn’t really have a public-facing vector; it’s all about sharing with your friends.
It’s entirely possible that what we need here is two separate things: a social media “Twinstergram” personal app and an RSS reader for public-facing content, no matter how short or long. (Time to pour another one out for Google Reader, the demolition of which was one of the most gratuitous acts of vandalism in the history of the Beast of Mountain View.) The RSS stream can be for interests and new discoveries; the Twinstergram can be about maintaining the close relationships you already have. It’s possible that something like micro.blog is the tool for that RSS, because it essentially puts the form of a social media stream on individual blog feeds – but of necessity, it’s based on public blogs, not private sharing. It gives blogs the utility of Twitter, rather than actually replacing Twitter. But let’s hang onto that.
I suppose in my life, the personal app would have to have circles for my self-selected family, for my friends in DC, the Bay Area and Nashville, maybe (if the architecture permits) for Vanderbilt stuff or tech stuff. And then, when you want to post something, you tick off what circles you want to be able to see it. So if I have a follower X, I may put X into circles for Bay Area and Nashville and Vanderbilt but not for tech or DC. That’s the difference: Twitter, as far as I know, lets you chop up lists of your followers for who you want to see about what, but not to whom you want to send. If you want to do that, you have to maintain multiple Twitter accounts, which people may or may not know you have, and have separate ones for friends, for sports and for technology, just for starters. In your life, you are many different people to the rest of the world, and as far as Twitter is concerned, each of them has to be a separate account, the end.
I keep thinking back to Path, which was another ex-Facebook spinoff in the era of “all God’s children gotta start an iPhone social app with photo filters” around 2010. Their big solution was to have Dunbar’s number as a cap on how many friends you could have – there was definitely a presumption that this was for keeping in touch with friends, not brands. Path had some of the Foursquare sharing built in to go with its picture filters, wakeup/sleep count, weather status and music logging. Cocoon seems to have a lot of the same built in, albeit with a 20 person limit for each group – and, of course, it’s still not cross-platform. The interesting shit still always seems to be on iOS first. Which in turn begs the question of whether Apple might not be able to put something together with iMessage, Find My, iCloud Photos and maybe even Weather and Music to create some sort of presence-sharing app that comes with security and confidence…but then, you’re leaving out 80% of the world. But after Facebook and (waves all around), you can’t palm off security as a consideration for something this personal.
The other thing that’s out there, of course, is the whole “roll your own!” of Mastodon and federated instances. All I can say about that is I try logging into my mastodon.social instance about once a year and it’s just an utter !-ing disaster area. Absolute Wild West lawlessness, like Usenet without the structure and authority. And again, the presumption that everything you post is public. Trying to roll your own solution in Mastodon feels a bit like moving twenty miles outside the city walls and saying you’re going to build your own commune there. Mastodon, like Diaspora before it, doesn’t pass the Ed Earl Brown test. You have to get everyone in one spot, it has to be trustworthy, and it has to be private.
The wild card in all this is RCS. Rich Chat Services were supposed to be the thing that brought the advanced features of WhatsApp or iMessage to ordinary text messaging, independent of any particular provider. The fact that the current push is entirely driven by Google’s effort to finally have a working chat app of their own suggests that carrier/platform independence may not be feasible, and there’s also the small matter of encryption – which is table stakes for any viable chat model. Nevertheless, if there were some way to couple public key encryption to RCS and make it platform-independent, that would be the framework on which a bigger tool could be built to organize multiple RCS messages into a sort of social media platform…and there you have it. Like micro.blog, what you need isn’t necessarily a whole social media application, but a tool that will put the appearance of such an app around a series of RCS messages, individual or group, and let you more easily keep them straight and share content among them. It only works if you have encryption built into RCS, but once you do, you can have an app manage that framework without ever needing to see the content itself – arguably doing its work entirely on the device itself. Just as Signal on Android can be your SMS app in addition to secure chat, this app – which, because I am a clever dick, will be called Venn – merely keeps track of your circles as a means by which to send blast RCS messages. The effect is similar to the earliest days of Twitter, which at its dawn was essentially a service for mass-texting a group of subscribers from your primitive 2007 feature phone. You’re sending a message to the group chat, but Venn is basically keeping your lists for you and providing a one-touch way to paste in location data, local weather, or what song you’re currently booming on Spotify.
This would require a lot. It would require RCS to have a viable encryption model, and it would require Apple to go along with it (right now the iPhone does not support RCS and there are no plans for it to do so, given that everything RCS promises has existed for years with encryption in iMessages). And two years after Google decided that RCS was its messaging future, it’s still only really a thing if you use Android Messenger, with no carrier in the US but T-Mobile able to interoperate with Google’s own RCS servers – and forget about an iPhone.
But Venn, at least, is a first cut at how you might go about replacing the most harmful creations Silly Con Valley ever barfed up with something at a slightly more human scale.