I used to read a lot of short news stories, especially during dead moments when reaching for my phone was the habitual action to take. This felt useful in the moment as I learned something about the state of the world that I didn't know yet, which pretty much is the definition of news. But I concluded that this wasn't a great use of time. So I decided to end it, which turned out to be a journey in which I ended up setting up my own VPN. Here's the how and why of that journey.

Background articles, podcasts and blog posts had always seemed more interesting. Nevertheless, regular quick news appeared a useful filler for those 2-minute gaps in the day where you're queuing or waiting for the microwave to beep. Often though, I would find that most such news is not newsworthy, or at least not above my threshold for newsworthiness, for a couple of reasons.

In the first place because for any single "newsworthy" event that occurs in the world, there are a dozen news articles about it over time describing the progression, lobbies, or aftermath. If you would've just read a summary article though, you would have learned almost the same. There are some obvious examples such as news about negotiations or elections. But even for events that do seem time-critical to read about, e.g. the spread of COVID-19 in your area, I'd argue that for most people it wouldn't make any material difference to their life if they instead read an update every couple of days (apart from new rules being announced, but there are other sources for that).

Secondly, many news articles report on events that you probably would not remember anymore in 2 years from now and do not affect your life in the foreseeable future. Is it really worth spending any time on that?

Even if you manage to distill the stream of news to only selected articles on topics that matter, you will still risk getting a distorted and more pessimistic view of the world, which Factfulness explains in an excellent way (and which I would highly recommend). Enough reasons to cut the habit.

However, I found that merely observing the above wasn't effective; it just made it feel more of a waste of time after reading another article. I tried iOS's screen time feature and it did help for a bit, but I quickly found entering the code to bypass the lock becoming habitual, at which point the effect was gone (I've heard people make other people set this code for them, which sounds like a good strategy as well). This mostly seemed a result of the fact that I could enter the code in under a second and hence it was very quick to bypass.

Another problem with this approach was that it didn't work well across devices. I would still be able to read on Chrome on my laptop for example. So I wanted to find a way to cut this habit that satisfied two main requirements:

1. It works across all my devices and browsers.
2. Bypassing it should be cumbersome enough that I am immediately aware that I'm bypassing it.

It got me wondering whether an effective way would be to insert delays into the websites that I didn't want to visit. That certainly would make me aware after wondering for a second why the page was so slow. It also most likely would mean I would need to leave my browser to turn it off, which seemed like a high enough hurdle.

So I decided to try this. The best way to get this to work across devices and browsers seemed to be to tunnel all my internet traffic through a single server and add latencies there. A personal VPN is ideal for this, so I forked an open-source project called Algo, which allows you to set up your own VPN using any popular cloud provider.

Algo made setting this up easy, and now I had a server in Amazon's cloud that had all my traffic go through it, from all my devices. The only remaining thing was to add the delays. After some searching, I found that most Linux distributions come with a tool called tc (short for traffic control) that allows you to throttle traffic at the packet level. A couple of hours of bash scripting later all my previously go-to news websites behaved like I was on a 56k dial-up connection (unfortunately it didn't have the appropriate sound effects...).

Surprisingly, after I spent a morning setting this up, I didn't have any urge anymore to even attempt to visit any of these websites. The simple act of consciously investing time into breaking the habit seemed to have been enough to cut it. It did still turn out to be useful a couple of weeks later though when I found myself occasionally trying to access them again, and then the page taking 10 seconds to load was a great reminder. By now, the habit is gone and I never visit those websites anymore.

In the end, I realised this was a similar effect to how paying for a service (e.g. the gym) can reinforce a positive habit: you feel you have invested in something, which raises the barrier to break it as you don't want to feel like you lost money. In my case, I wouldn't want to turn off something I have invested a couple of hours in to build. The delay itself merely acts as a reminder of that. Investing time into building your own solutions to end the habits you don't want to have helps.

You can find this project on my Github if you're interested in trying this out for yourself. Besides the delays, you also get (thanks to Algo!) DNS-level ad-blocking across all your devices and the privacy benefits of a VPN. It also works for mobile apps, although you need to figure out which domains the app connects to.