Roy Tang

Programmer, engineer, scientist, critic, gamer, dreamer, and kid-at-heart.

Blog Notes Photos Links Archives About

Noisy channels

There’s a significant risk of information overload nowadays. For someone like me who spends a lot of time on the internet, there’s a lot of feeds I follow. Not just social media like Facebook and Twitter, I also follow a set of RSS Feeds via Inoreader, I’ve signed up for a few newsletters, and on weekends I read through several Flipboard categories. Just another symptom of my tendency to want everything I guess.

It is arguably all too much. So I thought about Marie Kondo-ing some of that stuff. The main problem right now is my RSS feeds - I follow some feeds that simply have too much content, and this means I accumulate a significantly high unread count in the app, which gives me some anxiety. While those feeds have the occasional interesting post or article I want to read, the sheer number of posts means the signal-to-noise ratio is very low for those channels. So I’m going to cut the most prolific offenders and see if the content I would’ve gotten from there would surface some other way.

I do still like those channels, so I’ll also promote them a bit:

  • Hacker News - there is a lot of high quality content here, and many of them won’t surface through other channels. But there are a lot of posts, so the volume is still difficult to keep up with. I’ve instead moved them to a Flipboard section, and just check out that section from time to time.
  • Dev.to - less tech industry and more programming-related posts, many of them beginner-oriented so not really for me. There is the occasional gem, but the S/N ratio is worse than HN by far. It would be great to read all that content if your goal was to know everything in all areas of software development, but anyone who’s spent any nontrivial amount of time in the industry knows that’s a fool’s errand. I also follow them on Twitter, so maybe that’s enough instead of the RSS feed. They also have a weekly “most popular posts” post, I might look into whether I can somehow subscribe to only those posts and see if that’s sufficient
  • Slashdot - one of the oldest tech news blogs around, and I’ve been following them on and off for years. The content I get from their feed has a lot of overlap with Hacker News and the tech section on Flipboard, but occasionally they have an interesting post that none of my other sources provide. The volume of posts isn’t as high as the previous two, but the S/N ratio is worse. I’ll try living without them for a while.

Note that a feed having a low S/N ratio is in no way a condemnation of their content. It’s just that not all content is good for everybody, and hence some feeds are better for some people than others.

All of this applies to social media too of course. I try to unfollow ruthlessly. On Facebook, which I mostly restrict to friends and family, I unfollow anyone who posts content I dislike (conspiracy theories, chain messages, “let’s see who will read this”, inspirational messages, etc) and generally favor anyone who posts puns lol (I find the overlap between those two sets is low). Anyway, they’re still “friends” on Facebook, so they can still message me and stuff. On Twitter, every so often I’ll unfollow a bunch and follow some other set, change it up a bit. You need to adjust your feeds when it feels like you’re seeing too much of one type of content or too much you dislike.

It would be great to have some sort of machine learning that I can teach as to which kinds of posts I prefer and which ones I want to ignore. Supposedly, social media’s “algorithmic timelines” are supposed to improve the S/N ratio of your feeds, but my problem with those is that items aren’t being displayed chronologically, which makes a feed work. Maybe I should write my own feed reader! (Not the first time I’ve had this thought!)

Posted by under post at #Tech life
Also on: twitter / 0 / 681 words