Modifying your behaviour with a simple Twitter bot
About two weeks ago I created a simple twitter bot: @timetostretch. Stretching every two hours or so when you work at a computer has clear benefits. I wanted a simple tool to remind me to do this. Playing around with the Twitter API was a really quick way to build such a tool. In addition - I was interested in testing the idea of twitted reminders - if effective perhaps I would build a more complex and ambitious system…
Thanks to Grackle and Ruby building the bot was dead simple. I wanted the reminders to tweet me a video link to different sets of stretching excerices… and finding good stretching videos on youtube is by far what took the longest.
The bot quickly proved completly ineffective in modifying my behaviour. The reasons for this were:
- A lack of clear conscious descision on my part to actually stretch at regular interval. Taking a break and stretching remains a thing I think it would be “nice” to do rather than something I am actually committed to doing.
- There was no sound to act as an alarm / audio cue to stretch.
- The reminder to stretch was drowned in the flow of tweets. It was too easy to ignore the call to stretch if a more interesting tweet drew your attention. In fact overall random web surfing was a more likely outcome than stretching
- Despite that the bot used Twitter, there was no social aspect to the activity of stretching. I failed to build a support group that would share and support my goal of stretching at regular intervals throughout the work day.
What I’ve learnt:
- The merits of a quick “sketch” of a simple idea. Building a rough unpolished prototype allows you to test the validity of an idea and allows you to: “Fail early. Fail fast.”
- I’d like to code more fun stuff that don’t neccerally have to result in an amazing finished product. Experiment, try new ideas…
