… and coming too close to burnout.
The last time I wrote a blog post about what’s going on by now lies quite a bit in the past. Therefore it is time to reflect on what has happened since. What I wrote in the “escape” and “future” documents still kind-of holds, but the hatches and hooks I hoped to be able to rely upon didn’t hold what I expected. Almost every anchor I relied upon has failed, and I don’t want to overload the remaining ones.
With few exceptions, my environment in Chemnitz is in bad shape, it actually has gotten even worse since the last time I wrote about it. I want to avoid having to go back there at all costs. The previous blog posts detail a lot of what’s bad there.
Given the rise of LLMs has in large parts ruined the job markets for most programming and many mathematics jobs (if one wants to have the freedom to decide not use that stuff at all), I tried to basically “bury myself” in academia, unfortunately, academia, with some delay, is now also submerged deep in the LLM-soup. Almost all of the students constantly seem to use LLMs, and talk about it regularly, supervisors/PhD students are also regularly interacting with that stuff, and by now, even the lecture materials are getting generated by LLMs. Which means it doesn’t really make sense to go to university if everyone is just using LLMs for everything, when I want to avoid LLMs as much as possible, and instead, I’ll quit university for a while and just try to get a library pass instead, which is about as useful, it seems for now at least, but cheaper / less implied effort.
I can’t even escape that fully in the NLnet-sponsored Topola project, given the lead developer also occasionally uses LLMs, and everything is a bit chaotic.
Software development in general is currently seemingly far away from being fun because one constantly encounters “paper cuts”, dependency hell, etc.. And the proliferation of “vibe coding” has made everything that was annoying or bad about software development so much worse. Proprietary software is nowadays very untrustworthy by default (because it’s likely “vibecoded”), and an extremely large subset of open source projects have fallen to the same infection.
It doesn’t really make sense to stay in Aachen if I’m not studying at university, and can’t get a job here. I don’t want to stay in Germany because politics is accelerating towards criminalizing anything leftist and/or anarchist, and I want to be long out of here before 2029 fascism will probably rise again in Germany, either under a government of the AfD or the CDU or both… It seems very unlikely that politics will change course until then, and I don’t want to put massive effort again into moving inside Germany only to having to abandon that again shortly afterwards.
Thus currently, my default plan is to move to θεσσαλονίκη (Greece) with the money I’ll get from NLnet projects, try to get a random job there and relax… and attempt to get as much distance as possible from German politics, LLMs and most software development. I want to maximize the time I spend with learning and researching mathematics, type theory, control theory, and various abstractions, etc. and learning Greek. These were the only activities that were truly enjoyable in the past years, besides spending time with friends. And given that my friends regularly travel anyways, it doesn’t make too much of a difference where I live exactly as long as it is reachable by usual means of travel (i.e. trains, ferries and planes).
In the past weeks, I occasionally pondered if I could maybe still push through and try to continue studying at the RWTH Aachen University, but ultimately, I believe it is a better idea to quit for a year or so to not over-exert myself. I’m extremely exhausted from the past 2 decades.
I also tried to maybe fork some important software where the upstream has become sloppified, but for anything that isn’t tiny and stable, it feels almost impossible to pull off.