Blog

In data science and engineering, technical skills are often the quickest way to early-career advancement – a new contributor who can write decent code is immediately an asset, in a way that a business-savvy developer with no coding ability is not. Companies often publicly espouse the benefits of softer skills, but that may not ring true in entry-level positions. In some technical fields, non-technical strengths can actually be viewed as negatives by peers, a betrayal of the ethos of the programmer.

With the announcement of Apple One in September, I decided to reevaluate my loyalty to Spotify. If Apple was going to offer me a (slightly) discounted rate if I went all-in on their services, I wanted to have a good look at Apple Music and see if it could replace Spotify for me. After almost three months of use, my conclusion is that I could live with Apple Music, but I’d happily pay a small premium for Spotify instead.

I recently discovered and binged the Ben, Ben and Blue podcast. The hosts all work in STEM education to some degree, and while that topic isn’t explicitly the focus of the show, it comes up a lot. In one episode, Grant Sanderson (one of the hosts) talks about how it’s tempting to teach new ideas by recapping the actual discovery of the concept as it happened. For example, I don’t know the exact details of the creation of the internet, but you could imagine an explanation that walks through the journey of universities wanting some kind of decentralized communication protocol, and how that effort was funded by government money with the justification that it could be useful for national defense.

For those not already in-the-know about Tony Horton’s infamous X-treme workout series, P90X was a collection of video workouts released in 2005. They’ve achieved something of a cult following in the years since, and Tony has returned in several reprisals – though none seem to have gained the fame of the original P90X. The workouts require minimal equipment: just a pull-up bar, a yoga mat, and a modest collection of dumbbells.

I spend a lot of time on meta-productivity: thinking about how to get work done better and faster. The majority of that thinking is around technology and specifically the tech tools I use. Maybe thinking about my workflow has returns large enough to offset the time it consumes in my life, but frankly it doesn’t matter too much to me since life optimization is as much a hobby as an objective.

For a while now, I’ve been thinking about how interesting it would be to look at data on my running habits. Depending on the time of year and some other factors, I probably run 1-4 times a week, and I’ve been running semi-regularly since college.

For the vast majority of that time, I’ve recorded my runs in Under Armour’s MapMyRun app. It’s a bit buggy and feels non-native, but it has a good Apple Watch app and collects the information I need fairly reliably.

Over the past few years, I have heard many people repeatedly note the downsides of Twitter, particularly its impact on our society. I don’t agree with all of those criticisms, but some are legitimate and most deserve a serious discussion. This isn’t that discussion, though. Instead, I’d like to spend some time on the many merits of Twitter, so lost in our current debate. Good Writing Or: If I had more time, I would have written you a shorter tweet

My colleague Brad Boehmke and I made a guest appearance on the Talk Python to Me podcast to talk about the challenges of supporting a large data science department that uses both R and Python. You can listen to the episode on any podcast player.

I was recently interviewed for 84.51˚’s podcast on the topic of upskilling the data science function.

I contributed a walkthrough to some basic Linux utilities for data scientists to the 84.51˚ tech blog.