I accidentally built a serendipity machine

3 March 2019

I love using Pocket - the service once known as 'Read it Later' and now owned by Mozilla. Pretty much anything I see on social media or the web that looks like it might be interesting and take longer than 30 seconds to read goes into my Pocket account. Taking advantage of the Pocket API, I've also integrated it into Aus GLAM Blogs, experimented with doing the same thing for any RSS feed, and created a script to deduplicate items in a Pocket list.

My latest Pocket project is called pocket-snack and was born out of a conversation I had with my comrade and fellow Pocket-lover lissertations. We both faced the same dilemma a gore-loving Netflix account holder has: overwhelmed by so much choice, it's difficult to choose any one item. So our Pockets continued to fill up, the dread of opening them grew, and the anxiety caused by all that unread material hovered. This is a familiar problem for those with over-large physical 'to be read' piles (I have one of those too), or sheds full of junk that 'might be useful one day'.

I came up with a solution using a Python script that tags everything in the list with tbr, archives the lot, and then re-adds just a small number of randomly-chosen items back into the list. Now instead of having a list of literally hundreds of unread articles from which to choose, I have a dozen or less: a sensible number that can easily be read, or at least processed. The list is refreshed daily, weekly, or on demand. Two things became evident once I started using pocket-snack:

  1. Dealing with a large group of articles by chunking it into smaller groups is surprisingly effective both at getting any traction at all, and significantly speeding up the process. It forces you to focus on just what is in front of you. I feel that this has also helped me to focus on the thing I'm reading - I've tricked my brain into thinking it's only one of 8 items rather than one of 308, so there's no need to rush or be thinking about all the others.

  2. Randomly choosing from a large list of articles I have consciously bookmarked for future reading over the last several months sometimes creates serendipitous sets, or serendipitous timing. Things I bumped into online months apart but on the same topic will sometimes appear in the same 'snack'. At other times, I've been talking with friends or colleagues about a topic and a relevant article I'd forgotten about will appear. I didn't mean to make a little serendipity machine, but it seems that whilst I was just trying to make something to keep my brain a bit quieter, that's exactly what I built.

You might call it serendipity.

If you have a Mac or Linux machine with Python3 you can download the latest release of pocket-snack and use it yourself.