Sway News


I’ve been working with Mike for a few years now on Sway. In 2016 in was conceived as a way to introduce people keen to broaden their reading on a political subject of the day, in return for a weightier vote in an opinion poll on said subject. There was a whole reading+voting app for that, and people tried it out. We learnt some things, but it hasn’t saved democracy or turned the tide of shit.

We’ve continued noodling on and off with it for a while since then. Here are some barely structured thoughts here to navigate to where we’ve got to.

Part of the problem we were attempting to solve was that not all news sources are good and on social media some news sources are actively bad or worse. Technically it’s possible to rank, check and flag content. So how about some way to see that ranking and flagging?

Another part of the problem is who is sharing these news stories with you. Within your direct network, you hopefully know and trust some of these people, but in the land of forwarded links, excitement can run ahead of fact and balance.

So how about some way to map the people outside your network to see what else they think, what else they share, what else they say? Sentiment analyis plus other behaviour should be fairly straightforward - it’s time consuming for a person to check out someone’s history, but programatically it would be instant-ish.

We had been thinking about moving the activity into a browser extension. Like Grammarly checks what you write, it would check what you see and provide:

  • background checks into who promotes what you are reading
  • where what you are reading came from
  • what other views that source holds
  • suggestions for better/balanced/deeper things to read on the same subject
  • keeps an eye on if you read the article before blindly forwarding
  • usefully, it could keep a history of things you read
  • in more depth, do a self-analytics report, “Strava for social media”

Once you let the browser extension strategy do some of the heavy lifting you can overlay memberships and other networks on-top of a broader one, for example: A small media brand has a subscription membership model, their community is based on being an informed bunch of likeminded people - you could badge these people in your feed, to more easily see ‘your tribe’.

While the browser extension model is potentially powerful, it has limitations. Pushing further in this direction, would it make sense to reach a bit further and connect the people and content on a separate page or feed? This makes it a “new” social network of it’s own. Open questions on that include:

  • is there enough glue to connect these people on a separate site?
  • how do you get there?
  • what’s the discovery mechanism for people to filter content + users on it?
  • what’s the user benefit, if the browser extension features are already working?
  • is there significant additional technical infrastructure to run it?
  • what does joining look like? How do you leave?
  • how does it integrate with the other dark spaces, messaging, private groups, etc?