What if social media optimized for less time online?
The strange thing about social media is that it borrows the language of friendship while measuring something else entirely.
Community, connection, sharing, belonging. These are the words on the tin. But the machinery underneath is usually pointed in a different direction: refresh, scroll, react, post, return. The platform succeeds when you stay. It succeeds when you check again. It succeeds when a small uncertainty has been placed in your mind and the easiest way to resolve it is to open the app.
This is not an accident, or at least not anymore. The large social platforms are attention businesses. Whether the revenue comes from advertising, subscriptions, recommendations, paid boosts, creator tools, or some softened combination of all of them, the underlying physics are similar. More time on site creates more chances to sell something, measure something, rank something, recommend something, or persuade someone to come back tomorrow.
The same gravity appears even in platforms that present themselves as calmer alternatives.
A writing platform becomes a feed. A feed becomes a performance layer. The performance layer becomes a recommendation engine. The recommendation engine starts nudging the writing itself. Before long you are not only publishing thoughts, you are composing them inside a room whose walls are quietly teaching you what travels.
I noticed this most clearly after spending time on Substack. I had joined for pragmatic reasons. I wanted somewhere to write, somewhere with an audience already present, somewhere with enough feedback that the habit might stick. For a while it worked. I found good writing there. I found people worth reading. I found enough signal to convince myself the arrangement was healthy.
But Substack is a social media platform that has convinced many of its writers it is not one.
The notes feed, the recommendation engine, the nudges toward engagement, the comment games, the growth advice, the small dopamine puzzles dressed up as community. All of it serves the same old metric: time on site. The fact that the money comes from subscriptions rather than third party ads changes the shape of the incentive, but not its direction. They still need you there. They still need writers to become more visible, readers to subscribe to more writers, and everyone to keep circulating through the platform.
Two forces seem especially corrosive: short form content and the algorithm.
Short form content is not inherently bad. A sentence can be beautiful. A note can be useful. A joke can reveal something true. But at platform scale, short form content tends to reward immediacy over thought. It rewards the line that can be understood without context, the opinion that can be agreed with or attacked quickly, the memorable phrase that feels like insight because it moves well.
The algorithm then turns those rewards into instruction.
You learn what works. Nobody has to tell you. The platform simply shows you. A divisive premise moves faster than a careful one. A tidy conclusion does better than an honest ambiguity. A headline with a clean angle travels further than a piece that refuses to collapse itself into a lesson. Over time, without anyone forcing you, you write less of the second kind and more of the first.
That is the algorithm working as intended.
The same pattern appears elsewhere. I saw it when looking at photography content. Photography is, to me, a slow medium. Images take time to make and should be given time in return. Yet so much of the surrounding online culture has become indistinguishable from marketing. Gear reviews, lifestyle promises, secret formulas, “the mistake I made for ten yearsâ€, “don’t buy this until you watchâ€, “how I finally learned to be authenticâ€. Sometimes the product is a camera. Sometimes the product is a workshop. Sometimes the product is the person themselves.
This is social marketing: the conversion of human expression into material optimized for circulation.
I do not think the people participating in it are uniquely bad. Most are responding to the incentives in front of them. If the room rewards a certain kind of performance, the room will get more of that performance. If the platform rewards frequency, people will post more often. If it rewards outrage, they will sharpen their edges. If it rewards intimacy, they will package intimacy. If it rewards advice, everything becomes advice.
At some point I began to wonder whether the problem was not simply bad social media, but the assumption that social software should optimize for being online at all.
What would it mean to design in the opposite direction?
Not a platform that helps you spend more time with strangers on the internet. Not a feed that turns every private interest into public performance. Not a popularity contest with profiles, follower counts, streaks, recommendations, and little numerical shadows attached to every social gesture.
What if the goal of social software was to become unnecessary as quickly as possible?
That is the premise behind KILTA.
KILTA is an experiment in local social software whose success case is not engagement. The success case is that you close the tab, leave the house, and meet people.
The idea is simple. You add the things you care about. Not to build a personal brand, not to gather followers, not to perform a version of yourself for a public feed. You add them so that, if enough people nearby share enough overlap, the software can help turn that overlap into a real-world gathering.
That might be bouldering, sketching, board games, repair cafes, vegan meals, reading groups, mutual aid, walking, environmental projects, learning a language, making music, or meeting a few people in a public place who care about the same small thing. The subject almost does not matter. What matters is that the internet is being used as a bridge rather than a destination.
This changes the design constraints.
If the goal is to get people offline, then a feed is suspect. A feed is very good at creating the feeling that something is happening, somewhere, just beyond the current edge of the page. It is less good at helping five people in the same neighbourhood arrange to do something useful next Thursday.
If the goal is real-world community, follower counts are suspect. They import status games before the community has even begun. They encourage people to optimize for being seen rather than being present.
If the goal is thoughtful connection, short form performance is suspect. Not because brevity is evil, but because performative brevity becomes a game very quickly.
If the goal is local gathering, global virality is suspect. A million users spread evenly across the world may be less useful than a hundred people in one city who all want to do the same thing this weekend.
So KILTA has to be quieter than the systems we are used to. More anonymous in some places, more careful in others. Less interested in broadcasting identity, more interested in finding shared intent. It has to treat attention as something borrowed from the user, not something captured from them.
There are hard problems here.
Anonymity can protect people from performance, but it can also protect bad behaviour. Locality makes gatherings possible, but it also makes safety more important. A network like this is not useful until there is enough density, but chasing density too aggressively risks recreating the same growth incentives that made the larger platforms so corrosive. If the software helps people meet, then moderation cannot be an afterthought. Nor can spam, harassment, creepy use cases, or the ordinary awkwardness of strangers trying to coordinate in good faith.
I do not have all of this solved. That is partly why I am building it.
But I think the direction matters.
For the last twenty years, much of the web has been shaped by a small number of very large companies trying to become the layer through which everything else passes. The platform you use to write, the platform you use to talk, the platform you use to publish photos, the platform you use to find events, the platform you use to hear from friends. Each begins as a convenience. Each slowly becomes an environment. Then the environment starts making decisions on your behalf.
The alternative does not need to be grand. I am increasingly drawn to small, human-scale things. Tools made by people who use them. Services with constraints. Places that do not need to swallow the world to justify their existence. A thousand smaller trees rather than a handful large enough to shade out everything else.
KILTA is one of those trees.
It is brand new, which means it will be quiet at first. That is the honest state of any network before the network exists. In a big city, it may become useful sooner. In a smaller town, it may take longer. The cold start problem is real. But the way a thing grows has consequences. A social network grown through outrage, vanity, and algorithmic reward will carry those patterns in its bones. A social network grown through small groups of people trying to meet in the real world might carry something else.
The internet is not the enemy. I do not want less internet because I dislike technology. I want better technology because I still think the internet can help people find each other.
But finding each other should not always mean staying online together.
Sometimes the best thing software can do is introduce the right people, make the first step easier, and then get out of the way.
Sounds interesting? check out kilta.io