Quantified Self and Privacy
Published by James Gallagher on .
This article takes approximately 5 minutes to read.
I am somewhat hyper due to the coffee that I have just consumed. I shall leave describing said coffee for a future post.
Over the last day or so, I have been frustrated by whether I should invest my time in quantified self projects when privacy is on the line. The more data that I collect, the more data there is that falls into the hands of other companies. My coffee tracker has been hosted on Airtable. I have ten services which are integrated to my Exist.io account.
I was listening to another episode of the Mozilla IRL podcast earlier today. The episode I listened to was on privacy. The person being interviewed made a point that has been stuck in my mind since I heard it: you should be extra careful when deciding what to put online because you never know who you will be in the future. This is the problem I am facing with quantified self. Should I collect so much data when every additional data point opens me up to a greater risk of my privacy being violated?
The Question of Privacy
I believe that privacy comes down to my personal parameters. What is it that I am willing to accept?
I feel comfortable letting companies like Fitbit analyze my data, at least for now. This is not to say that in the future I will not stop gathering data. I believe that, where I am in life, it is acceptable to let the companies that I use for quantified data gathering store my data. It’s either I let Fitbit store my data or I do not get to collect it at all. This would mean that I could not continue with my quantified self work.
My policy on privacy is not blanket. This is something I am realizing as I think about what services I should, and should not, use. Although it’s somewhat scary to think about what could happen if my Fitbit data was leaked, if I applied this principle to every service then I would not use the internet. I do think that it’s okay to share data with some companies because they provide a valuable service. With other companies, I do not have that level of comfort.
Despite using Google services for work, I still use ProtonMail. ProtonMail has a bit more friction. I need to enter my password and get a 2 factor authentication code every time I want to view my emails from the website. I am comfortable with this trade-off for privacy because I believe that my emails need to be kept secure. I don’t really use emails much; mostly just to sign up to services and talk with a few friends. My mind has slowly adjusted to becoming used to more privacy for emails. I feel better about using ProtonMail. That’s why I pay for their services.
Quantified Self Projects
The position I am in right now is that I want to collect data but I’m not sure if I want to share it. That is why I am not presently working on any public analysis tools. I did make some data available on my site earlier this week. That data has been removed because I did not finish implementing the structure of that data. I do plan to revisit this at some point in the future. I just need time to consider my next steps.
There are two reasons why I am so interested in quantified self. For one, I love playing around with data. There is something really exciting about running numbers in a spreadsheet (or, even better, in a Python program) and using those numbers to derive insights about my life. I do plan on doing a bit more financial analysis. I think that will be really fun. I hope to do this on Kaggle so that I can refine my programming skills at the same time.
Quantified self also lets me learn more about myself. What cafes do I visit most? What coffees do I drink the most? How many grilled cheese do I eat on average each day? These are all questions I can answer with my limited data sets that I’ve been working on over the last month. 1
I do need to sacrifice a bit of my privacy to track this data, but not too much. If I keep my insights to myself, most of my concerns go away. The world does not need to see what I learn about myself. Quantified self is an inherently personal hobby. I learned this after watching a few lectures on the Quantified Self website. The Quantified Self organization recommend that lecturers try not to generalize. This is because everyone has their own setups and goals and ideas. My setup will be different to that of everyone else because I am tracking other metrics.
For now, I am going to continue my quantified self work. I’m about to dive into my finances. I plan to keep my findings to myself in the interim. I just enjoy data collection. Like coffee, it is a hobby, and one that I don’t feel like I need to talk about every moment of the day or share on my website every time I make a small discovery.
This may be cheating but this is Day 2 of #30daysofdataownership. It’s not exactly a project but this is a question that has been burning on my mind and I feel like I needed to explore it in more depth before going any further on the #30daysofdataownership train.
Full disclosure: the grilled cheese data set is only a few days old. ↩