Leaving Academia

Posted on Fr 27 Januar 2023 in misc

This has been on my phone(!) far too long, so I might as well post it.

Before talking about why I left academia and how I feel about the transition, let me first talk about how I entered & worked in academia.

University was the first time I felt like I belonged. My biology teacher told me "once you are at university, you will meet people more intelligent than you" (meant as a promise, not a threat 🙂). He was right and I met quite a few people with shared interest.

Due to circumstances, I started specializing in NLP, was offered a fully funded PhD position where I could work on whatever I found interesting (double edged sword, but a different topic) and I enjoyed it. Getting there was a nice combination of the path of least resistance and my interests.

Working towards my PhD was quite stressful at times (this is probably true for everyone) and when I reflected on my situation, I came to the conclusion that I stressed myself out of fear of being bored; the thought of working in a non-demanding job where you just do routine stuff is stressful for me.

I then thought about what to do after my PhD and my wife and I agreed to test how staying in academia would work. This meant that I started commuting 1400km a week by train (700km and 8h each direction). I was only at home during the weekend. People often assume that this kind of traveling was taxing, but I found it to be quite relaxing. The taxing part is being away. I could not help in the household, bring our child to kindergarten and there is nearly no social interaction on either site. Sometimes I traveled through Germany, sat in an empty office, met people for an hour and then traveled back.

A career in academia usually involves having to move from university to university and to either take your family with you or not. It involves uncertainty on whether you will still be able to work in your current field in three years and, at the same time, your perceived value in industry does not increase, switching becomes harder. I have seen several people in my field applying for permanent positions for years, having grants falling through, moving to different cities, countries, continents for temporary contracts. I also saw how professors (which in Germany is mainly "full professor") have to do a a lot of admin work, stress over grant applications because they have a responsibility for their department, work during vacations (I did that as well, but they have tenure). All these cons require enough pro argument to make it worthwhile to stay in academia. I decided it was not.

My main reason for working in academia was that I could work on interesting problems, problems where one needs to talk with others to even understand it, where we then can focus on the core of it.

When reflecting on this pro/con, I came to the conclusion that I can have a better trade-off in industry, with my hypothesis being that academia is not the only place where interesting problems are being worked on. I applied at two companies and decided to start at New Work because the people there were closely aligned with what I was searching for. In the recruitment process I spoke with several people who left academia after a postdoc and who (I felt) completely understand my motivation to switch and my mindset.

When I decided to quit, my decision was validated by the discussion I had on whether I could take three out of the seven weeks of my accrued vacation time (I already proposed to not take all my time for a good hand-over). It was a nearly 1h discussion circling on whether I could guarantee a paper if I take three weeks of vacation over the summer. In contrast, at my current place, I get emails reminding me to please take vacation and the company actively promoting healthy work/life balance.

So, what does my work look like now and do I actually have interesting problems to solve?

First, the interaction with other like-minded people. We have more than 100 people in our internal data community. These are data analysts, data scientists, data engineers. Some of them work on problems I find interesting and some don't, but overall there are more people to interact with on a day by day basis than I ever had in academia. Most importantly, there is a much higher willingness to discuss problems and spend time on interactions that might help the other more than oneself. I think this is because the success measure is more applied on the company level than in the personal level ("I helped X write a paper" doesn't really count when applying for jobs in academia). There are also many more opportunities to interact with people who are excellent in completely different areas (marketing, UI, corp dev, ...) where I am just delighted to see their work as an outsider (and sometimes work with them).

Second, the tasks I work on. I primarily develop products, so there is no foundational research. Nonetheless, most of my training from academia is quite helpful: Formalizing problems (e.g. going from a user pain point to a training objective), structuring approaches, explaining to other DS, developer and also non-tech people (teaching pays off, finally!). I spend much more time talking and planning than in academia and if you know me you know that I like to talk.

The roles in industry (at least in my case) feel like they are more in flux and adaptable. One can change while being in a role, slowly grow to a different role, in contrast to the largely static work requirements unless switching jobs in academia. This organic growth means that changes are more aligned with the person, that there is a more organic growth in responsibilities, flexibility in moving positions (where & when). The reward function is much more gradual in industry and less binary than in academia -- no "paper published or not", "funding or not", "new position or not".

One of the most important aspects: My work is much more family friendly. The ability to allocate time relatively freely, being able to just stop working with good conscience (no paper deadlines, no assignments to grade over the weekend, …) and actually not thinking about work-related topics after work is huge.

So, overall: do I miss academia? absolutely, at least certain aspects. Was the move to industry the correct one? Again, absolutely.