Introduction

This will serve as both the first blog post on my site, and a way for me to explore how I feel about personal websites (and personal brands) for academics in general. For this, and all other posts, these views are of course entirely my own opinion and are not to be attributed to anyone I work for or with unless it makes them look better in some way. With that out of the way, let’s begin.

Gone are the days when you could march out of your PhD into a cushy assistant professorship, confident in your inevitable rise to the holy grail of tenured full professor. Now that there are way, way more new PhDs than there are faculty positions, we all must either head for the fabled greener grass of industry or fight over the scraps available to us in academia. For pretty much all fields, this means publish like mad. But in statistics, and specifically machine learning, I’ve noticed an additional requirement that has cropped up.

First, without ruffling too many feathers, I think it’s pretty safe to say that most of the basic ideas of machine learning were developed in statistics long before the computing power existed to implement them. On the other hand, a lot of the cool, practical breakthroughs today are by computer scientists. This has led to the field being almost entirely adopted by computer science, and so statisticians who want to work in this field have to market themselves the same way software developers do, since that’s who these companies are used to dealing with. Consequently, we find ourselves in the purgatory of personal projects.

It’s not that I don’t recognize the benefit of messing around with these tools in order to learn, but as someone who is interested in statistical theory, I wish there were more options for quick projects besides analyzing a dataset. As you can see from my projects page, I do have an example of such a project, but I was required to do that for class. On my own time, I have much more interest in reading and developing theory, which is quite a bit slower than a quick data analysis. All this is to say that I think companies need to realize it’s possible to be a self-motivated learner without having an app startup, and they could improve their hiring diversity by not placing so much emphasis on personal projects. However, I’ve yet to come up with a way to measure “mathematical curiosity” that is as easy to measure as number of projects for “coding curiosity”.

Advertisements