bug repellent

a blog by @captainsafia

<= go home

The data fiend’s dilemma

I’ve got a confession.

My addiction to data might have gone too far.

About 6 months ago, I became more focused on health and fitness and began to center my lifestyle around running. I began to eat better. I started to work out 6 times a week. I ran anywhere from 4 to 6 times a week.

And I began to obsess over measuring every little bit of it.

I tracked my runs, my workouts, what I ate, how much I drank, my body fat percentage, my weight, my muscle mass, changes in my menstrual cycle. Everything.

Data ruled by world. I worked out and ran, not just for my health, but to add another row of data to my spreadsheets and CSVs.

And I’m not sure that any of these numbers have changed anything — I’m not entirely convinced that the trajectory of my fitness journey has changed because of data.

And I think this is a part of a larger problem that exists in our IoT-centric, data-driven (should I add more buzzwords?) world that we live in: being data-driven doesn’t necessarily make us better. It makes us more informed in some ways, yes, absolutely, I agree.

But more informed isn’t necessarily better — especially when you have to fight the forces of bias, ignorance, and plain ol’ stubbornness.

This is the point in this post where I’m supposed to present an answer to the crisis of disinformation — the desire to disengage with the information that we have — but I don’t have one.

This is just something that I think about quite often, especially because it has affected me so strongly during my fitness journey.

And now that I think about it, although I cannot lay out a complete solutions, I can lay out a list of smaller steps we can take to addressing it.

  • Encouraging people to have a deeper awareness of how machine learning and data are used in their day-to-day lives. People here means all people.
  • Be honest (and humble) about where machine learning can’t help us and won’t ever be able to (at least anytime in the near future).
  • Stop making everything an IoT.

Rant over.