“Free speech is the bedrock of a functioning democracy, and Twitter is the digital town square where matters vital to the future of humanity are debated” is how Elon Musk explained why he spent $44 billion to acquire Twitter.
That phrase — “digital town square” — has a long and messy history in social networking. Facebook and others have proven that throwing millions or billions of people into a single unfettered space is a mostly impossible and mostly terrible idea. But maybe we’ve just been defining it wrong. In reality, a town square isn’t a place where everyone stands in a crowd and yells at each other while advertisers throw things at them. It’s a place where groups of people find each other and spend time together. And yes, maybe they debate matters that are vital to the future of humanity. Just not with everybody else.
Twitter shouldn’t try to optimize the public arena, a 200 million-person conversation that will never make sense. Instead, under Musk, the company should focus on the private side of the platform, a woefully underdeveloped system for messaging and communicating that could turn it into the best messaging app on the market.
Twitter should invest heavily in making DMs a powerful, searchable, encrypted messaging system. It should finally roll out the much-rumored “long tweets” feature that lets people post more than 280 characters. It should continue to work on the Communities feature so people can chat about things they’re interested in rather than junking up their followers’ timelines with stuff they won’t care about. It should integrate Revue newsletters and focus on making Spaces more useful and reliable. It should worry less about ranking your timeline and more about giving you ways to talk and people to talk to.
It’s true that Twitter is the internet’s best platform for building an audience and then sending that audience to other places. You build an audience on Twitter, the joke goes, and monetize it on Substack or YouTube or any number of other places. Twitter should keep leaning into that because it’s what gives the platform its cultural cachet — after all, what is cable news if not just a bunch of people in suits reading tweets?
But most people don’t want to communicate with 200-some-odd million people at a time, and even those who do don’t want to only do that. Instead of just focusing on connecting people to the public square, it needs to focus on connecting people to each other. Twitter should catch up to where Facebook and others have gone over the last few years: toward a more private version of the internet, where hanging out online involves less shouting your angriest thoughts into a teeming mass of thousands of equally angry strangers and more actually spending time with people you care about.
When Twitter rolled out Communities, David Regan, a product manager, wrote that “we haven’t done enough to help connect people who are into the same things.” This has been a long-standing problem for Twitter, which has tried Suggested Follow lists and Trending Topics and Fleets and a thousand other ways to give people stuff to do on the platform. Communities is the first correct approach, just giving people a safer, quieter place to be together. That’s why Facebook thinks Groups is the future of the platform; it’s also why Telegram has grown so fast and why WhatsApp launched Communities just to keep up.
There’s plenty of competition in the group chat space, but Twitter enters the fray with an advantage: communication happens on a sliding scale. Sometimes you want to text your best friend, other times your friend group, other times your whole company, other times the world. Most messaging platforms excel at one, maybe two of those things. Twitter could bring its fast, simple communication tools to the whole scale. And by putting it all in one place, in one app where users only need one username, it could be the best communications platform on the planet.
To be clear, none of this is easy! Many content moderation problems can be harder in these semi-private and private spaces, though Musk and Twitter could learn a lot from platforms like Reddit and Discord on that front. And for a platform as globally visible as Twitter, those problems will be magnified, especially as government regulations around the world get more stringent.
Before Musk came into the picture, though, this sliding scale approach appeared to be roughly the road Twitter was headed down. As long ago as 2016, then-CEO Jack Dorsey was saying that “at its core Twitter is public messaging,” and that speed and utility were more important to the platform than character limits or specific timelines. In practice, the company has long focused on the one-to-many messaging — to the detriment of everything else. Over the last couple of years, the company has, at long last, started to regularly ship products, and many of them were meant to make Twitter feel a little smaller for users. It launched Communities, through which you can tweet but only to a group of people interested in the same time. It started testing Flock, a way for you to just tweet to your close friends. And it finally, for the first time in forever, seemed to remember that DMs exist and start to roll out a few improvements.
Of course, in the long run, projects like Bluesky mean Twitter could be even bigger than Twitter. It could become a universal standard on which lots of different kinds of experiences could be built. For Musk, who periodically debated starting his own social network before deciding to just buy one, the idea of a decentralized social platform ought to be an enticing one.
In general, much of Musk’s focus on Twitter has been on the algorithm. The theory behind open-sourcing it, so far as I can tell, is to give users transparency and choice about what they see and where. But the better strategy is to let people build their own experiences, with better tools than just the follow button. Help them and let them figure out who they talk to, what about, and how, in as many ways as possible. That’s where people will talk, where they’ll feel free to actually freely express themselves. No algorithm is ever going to do better than that.
