The previous post has dealt with the old – more precisely the end of my first PhD attempt and what has resulted from it. This post deals with the new – my second PhD attempt, how I got this chance and what it will bring for me.
After I mustered up the courage to quit my old position in Bonn even before I’ve secured a new one, I started looking for jobs again. This time, I actually “inverted” my priorities somewhat, since I was quite put off by my old position. I deliberately searched for positions with applications in the natural and life sciences, since I have lost my fear of inter-disciplinary research in the meantime. Also, I looked out for smaller research groups which made a more sympathetic impression, as I realized that communication and getting along with my colleagues and my supervisor is more important to me than the group’s prestige or equipment. I sent out unsolicited applications to universities, research labs and start-ups, and notified my master thesis professor that I was looking for jobs again. In the end, the response from all the unsolicited applications has been quite underwhelming (if they responded at all), while through notifying my professor I actually secured my new job. He shared the email address of a leader of a research-oriented group inside the technology and innovation division at SAP SE, and from the start the atmosphere in my calls with him and his group has been very friendly and relaxed. It seemed as if I was trusted from the get-go, which goes to show you the power of being connected and recommended by familiar people! We sealed the deal, and after a month of pause (during which I could recover from the stress of Bonn on a beautiful vacation) I started my new job on May 1st.
For the new PhD position, my former master thesis supervisor will take on the role of academic supervisor. However, since I am employed by SAP this is actually an industry PhD! This type of PhD position is sometimes looked down on by university researchers, as they consider it a “PhD for the soulless” or a PhD for those that did not manage to stay in “proper” academia. I also was exposed to these opinions from time to time, so I was somewhat doubtful when the prospect of entering precisely such a position drew nearer. For somebody who sometimes dreams a bit too much about becoming the very best and saving the world with his work, ending up in “boring” industry surely sounded like settling for a plain and boring life. It took some time to shake off these naive and somewhat irrational beliefs, but I think coping with the failed first PhD attempt actually helped me to get my feet back on the ground and to tone down on the dramatisation of rather mundane career events. If you’re aware of your values and do an alignment check from time to time, you’ll probably be fine simply taking the next best opportunity that arises to you and enjoying life on your way. Concerning the topic of my new position, we aim to combine Multi-Agent Reinforcement Learning with Graph Neural Networks and apply the resulting insights to autonomous data center networks in some way. That’s quite a funky combination of domains, and time will tell what precisely will be the problem we’ll try to solve and the contributions we’re making (and whether it will all work out in the end). In this new job I will have to serve the two needs of both my academic group and my SAP group, and I’m genuinely looking forward to seeing how I fare in balancing my time between the various meetings, and my efforts between the roadmaps and goals of the two groups. Ironically, the fear of having to do precisely such a balancing act was one of the main reasons I decided to go to Bonn… well well well, how the turntables. Anyway: Holding that balance so far is actually quite fun, and I will probably write one or another blog entry about the adventures of this balancing act.
From my first few posts in this blog I think it became quite clear that I enjoy thinking about the “big picture” and the “broader impact” of what I’m doing, and that I’m concerned about doing something that has a deeper meaning. Most likely, I exaggerate in that regard and end up worrying too much about finding the perfect next step instead of just enjoying the journey. At least that’s what I started telling myself after I secured the new job, because now my research has to at least somewhat align with the interests of my employer in addition to my professor’s research domain. This company doesn’t sell guns nor drugs and commits itself to sustainability in businesses rather than exploiting the earth for its fossil fuels and precious metals. Yet, since it is a huge multi-national company with commercial interests, getting paid by them means that I won’t be working solely on attaining world peace or diverting the climate crisis. What I now tell myself is: while you should never give up on their beliefs and ideals toward a better world, it will only break your heart if you stubbornly remain idealistic about this when it comes to navigating the adult world. There simply aren’t enough sufficiently paying “super hero” jobs out there yet, especially since housework, nursing, social work etc. tends to not get paid very well. Most people will have to bring home some money for themselves and the family, and if your job partially aligns with your plans to save the world it’s actually a good step forward for you already. For me personally, the PhD position enables me to finally work on Graph Neural Networks and Reinforcement Learning – two research areas that I find super interesting because I see a lot of potential in them (some quick google searches will be enough for you to convince yourself of what these technologies are already capable of). The application to data center networks also holds huge potential for changes that I actually do care about, such as reducing data center energy consumption and increasing their operational efficiency. If my research results are sufficiently backed by theory, they could also be applied to other problem settings which would benefit from intelligent decentralised control, such as smart energy grids or robot swarm control. In general, I think that increasing the resilience, robustness and independence of grid systems by means of decentralisation/distribution and intelligent control could be one of the things we need to deal with the huge challenges we will be facing in the near future.
In hindsight, the unsuccessful attempt in Bonn might have been just what I needed after all to get my priorities organised and become aware of what really is important in my upcoming professional life. Not only am I quite happy with the values fit of my job, but it also feels like I am partially freed from a pathological urge to constantly “see the big picture”, which often enough led me to disrespect my past achievements and my present work. And so, all in all, I am looking forward to fun and exciting times!